Well, I felt guilt because I feel I didn't try hard enough on the previous project log, so I decided to finally make a drawing so I can visualize the final thing more quickly than on 3D.
The resolution got butched, but I hope it is still clear enough. After all, I didn't add everything all at once because I wanted to make things clear enough.
I was having a hard time trying to imagine a solution for the shoulders, because, you see, everything ocupies space in the real world, so a Stewart Platform "shirt" probably wouldn't fit in the shoulder comfortably. So instead, I made an "open" stewart platform which would probably allow the arm to fit through it and allow for its rotation without troubling user's movement. Also, I did the same mechanism for the hips.
Also, I believe the human sketch I've made as reference isn't quite proportional to the actual thing, so I couldn't fit the abdomen linear actuators for torso movement.
I will try again with some random picture I find online.
Bruh
I will try again tomorrow or later, for the viewer it won't make a difference, lol.
There, now my exoskeleton has an abdomen (it actually took my like 5 minutes or something, lol).
However, I'm kinda confused on what I should do for the hips/thighs model.
You see, normally robotic hips are 2 degrees of freedom instead of the 3, but sometimes the 3 degrees are done in a way it doesn't properly imitate the human body.
But if you pay proper attention to your own hips/thighs, you can spread your legs (this sounded really wrong) even when you are in the sitting position, that's why I thought on rotating the entire thing in the drawing bellow:
Yes dad, I'm drawing butts for scientific porpuses!
(also, I forgort to draw a basis in the middle of the coccyx (the butt) in order to support the Linear Actuators to allow for a better rotational motion)
Of course, you would need that rotating ring on the thighs for the 3º degree of freedom I talked about in the previous Project Log.
Also, the more I think about it, the less I like the linear actuators for the shoulders/hips on the exoskeleton.
I really don't know how I would properly control and position these precisely on the exoskeleton...
But dividing multiple axis of rotation to multiple types of actuators would make the entire thing more energy efficient... While on the other hand, using a single electric motor for every axis/degrees of rotation would be more simple to build and position, but would consume waaay more energy...
hummm...
Maybe if instead of making a linear actuator, I make a reciprocating linear actuator, like a piston engine.
It will be easier to control (I think) and easier to make (I think).
... Also, I wonder if it woulad be better to make these in the legs/arms, since the human body motion is cyclical, which is not that good for electric motors to simply change speed back and forth all the time...
Taking into consideration the increasing complexity of the project, I wonder if I should actually copy older types of exoskeleton.
For example, the Hardiman, an exoskeleton from the 50s-60s.
It was basically an "H" with each part just loosely following the human movement. I doubt they thought it was the best, it definitely feels like they tried to make a reallistic aproach to the problem, but still the project didn't went anywhere as far as I know.
I call it an "H" because if both arms were lifted up, it would really look like a "H". But with one arm lifted, it looks like an "h"... lol
........ Is that why it is called "Hardiman"?
bruh
As far as I heard about it, it had a bad habit of breaking people's arms... 👀
Calculating Gearbox:
Soooo... This time I will try to make this thing with more care, finishing the actuators/parts as good as I possible can.
My brain stopped releasing dopamin and I started to feel like this exoskeleton is as useless and overly complex project just like the mech itself.
Not to mention that I don't even know how to properly make the exosuit follow the human movements in time. This is a problem for even the most well known egineers that actually work on these useless pieces of crap...
Well, whatever, I will think better about this tomorrow...
Well, I woke up and I literally forgot the most important detail about using drone motors: they spin too fast to revert their action in time (I think).
Of course, I can just use the reciprocating linear actuator I showed earlier, but the problem would still be acceleration... (I think)
Or maybe not, this looks to accelerate and stop really fast, after all the drone motor doesn't have a lot of mass... 🤔
(as you can see, I'm not very smart)
I also forgot that even though every actuator would consume 300 watts, it doesn't mean every actuator would consume 300 watts all the time.
I forgot that my previous project log has the torque and wattage values all wrong...
I actually need 1386.5 watts (1.8 horsepower) per electric motor in order to lift 100 kg with the arms that have 30cm of length.
Basically, 30 rpm is the jumping/running speed, for walking I would need way less than that. Actually 600-942 watts at 15 rpm.
But again, why have an mech/exoskeleton if you won't put it to work at full power?
Anyway, I will base myself on this 2400 watts electric motor, there are 900 watts electric motors around the website, but these can't keep up on 900 watts for long.
There is also the issue of having to have the lower part 3 times as strong as the top half, but I will just ignore that for now, it was just a rule of tumb for the human body.
And even if I ignored that for the mech, it would still need 5 horsepower per joint. if there are 5 joint per limbs... You get the idea.
So, anyway, this motor will be running at 33099 rpm and 0.4 newton meter of torque, so, 1386.5 watts, or 1.8hp. So it would need 1103.3:1 reduction gear ratio.
Of course I won't go straight up to that reduction, since the last "gear" will be a chain drive with 15cm of diameter, thus, needing "only" 200 kg instead of 300 kg.
It would then need way less reduction than that per step, but in the whole, it would be more or less this amount.
Well, the good news is that the calculation for sprockets is the same as for gears, since the sprockets are technically just round gears.
And since the sprocket I downloaded has 30 teeth and minimum 10 teeth, this means that I will have a reduction ratio of 3:1 at the end. This means² that I would need 90 rpm at the sprocket that connects to the reduction box.
And thus, I would need a reduction ratio of 367.77:1 in the reduction box.
But since this number feels a little broken, I will round it up to 370:1, which will need 33300 rpm to reduce to 90 rpm (I think).
Actually, after putting numbers after numbers and calculating a lot, I think the best course of action would be to use stackeable planetary gears.
So, yeah, the number is quite broken, I would need 38880 rpm 2:1 reduction and then 3 reductions of 6:1 planetary gears (or this one), 4 stages in total (wihtout counting the chain drive).
(Just now I noticed that the first link has slightly curved gears and the second simply doesn't have downloable 3D files, so I guess I will have to modify this planetary gear. Which I was already planning on doing to the other reductions, since I intent on adding bearings to each gear, carrier and axis)
There is also this stackeable 4:1 planetary gear, which I believe I will be forced to use, because just now it came to my mind that maybe people on the internet doesn't make the most precise gears... And this guy seems quite competent, I believe.
The only bad side of this is that I would need around 46080 RPM with 256:1 ratio (4 stages) and the 2:1 at the end with the 3:1 chain drive.
(You have no idea how many hours it took to me to reach this conclusion with these ratios)
... Actually, I think I will mess with all of those until I find a model that is well suited for 2D printing and/or laser cutting.
if you assume best case scenario of 3% of loss on the 6:1, then you would have 12% loss, meaning that I would lose around 139.08045417 watts.
Meaning I would need 1361.5 watts in total per joint on the exoskeleton, with 43336 rpm on the input with a reduction ratio in total of 432:1 and 1296:1 including the sprocket.
Almost on the limits of rpm of the electric motor itself.
Well... I believe I would need a lot of 1:1 bevel gears, since putting this chonky gearbox directly on the joint would just be too bulky.
But if I put a 2:1 bevel gear at the end, I think it would be better...
(this is a bevek gear btw)
Well, I don't think I would be able to DIY this design, but I think I could 3d print this dang thing...
Or maybe I can...
I totally forgor that I already had the intent of copying pieces from already existing ones, the bevel gears are quite cheap, but I don't know if these small ones would be able to withstand the loads...
That's why I wanted planetary gears with straight teeth, because I can simply make them longer and/or wider in order to withstand the loads.
3D modelling:
Well... After rambling so much, I think I finally can put the 3D model together.
This abomination you're looking at is the "actuator box", I still didn't put a pretty cape on top of that, neither I have made the holes to connect everything, simply because it will highly depend on the screws and bolts I will have available at the moment.
Plus, the position of the sprocket will depend on the part of the body I'm going to build, but I will just let this armature like this for now.
Also, just now I noticed that I forgot to increase the length/size of the gears and all that stuff...
Bruh, I'm starting to hate 3D modelling...
Edit²:
The last 15cm in diameter sprocket wheel doens't need to be there if you don't want, like I said before: you can treat the limb itself like a giant pulley/sprocket depending on the distance in which you attach the chain.
I say this because the wheel will be combersome while being attached to the arm, knee or feet, which can leave the mechanism exposed to outside debris and impacts.
By the way, I know this amount of gear reduction looks stupid, but let's just remember that you can make a gearbox lighter by making it out of other materials, such as UHMWPE.
While an electric motor can't be made to be lighter...
I mean, I'm using small drone motors, but I can't really find or build an electric motor that rotates at 30 rpm and has a crapton of torque...
The electric motors that I could find that doesn't need this crazy amount of reduction ratios are 40kg in weight.
One would be fine, but I need at least 10 of those...
The only way I could really get to this specific ratings without these stupidly oversized gearboxes would be through costumized purchases.
But as you can imagine...
I just woke up, and I think I misscalculated everything again... Because I think the final result of the gear reduction will have too much torque.
If you use a torque calculator and convert 200 kg to newtons (1961.33 newtons), you would only need 294 newton meters. Which would be 924 watts of power...
I keep messing the numbers for some reason, how can I get so wrong all the time?
Well, at least the electric motor can have its torque and rpm manipulated, so the the only thing will be changing will be the rpm, and thus, the actuator will work with higher voltage than amperage...
Anyway, I think this is fine for this Project Log, I don't think I will be able to 3D model anything for the last 2 oe 3 days, so I guess it will be better if I share my current progress and hear other people's opinions on the reduction box subject.
Edit¹:
I just saw this video where a guy makes graphene from eggs (or so called "blood graphene") and then makes a bioplastic that is stronger than steel. :|
The only problem is that it needs an hydraulic press to be made.
I think I should add the date of the day I publish the project log, I feel like it could be interesting for giving an idea on how long I've been working on something...
I'm publishing this on a wednesday, by the way, 2 days after I created the project log.
Anyway, there is no way around the problems when it comes to the Mech itself, it is just too dang expensive and too dang heavy.
I think I should (finally) be following other people's advice on the subject and start with the exoskeleton.
(Actually their advice was to start small and simple, not complex like an exoskeleton, but I will ignore that until it comes back to bite me, lol)
Joints and Actuators:
So, the first problem one can spot with an exoskeleton is the misalignment between the electric motors axis of rotation and the body axis of rotation.
Well, I couldn't find anything that illustrates that, but I will try to illustrate it on paint.
I hope it is clear from the image itself.
But anyway:
since the exoskeleton is outside of the human body (for obvious reasons), simply attaching the actuators at the same positions of the joints won't allow the exoskeleton to fully follow the human body's movements.
Of course, some parts of the body are kinda simple, like the knees and the elbows, since these joints are fairly simple.
But still, in some areas, like the shoulders, the torso and the hips, you will need to construct the exoskeleton like its axis of rotation is inside the body.
For example, using some kind of mechanism similar to this ring rotary platform in some areas.
Although I get a little concerned because it looks a bit too bulky and I can totaly see it rotating beyond the human limits and dislocating a bone or two, so I would suggest an "open ring configuration", if such thing exists.
Another way of getting around this is using linear actuators, but the conventional ones (like the screw type) have too much friction (or so I heard). But there are other ways around it:
This stewart platform uses the rotary action of electric motors to make a linear action.
Now, this one guy attached a rod to a timing belt and the movement of the timing belt propels the rod forward.
Although now I need to figure out the speed/rpm of the motor for the linear action and the linear action to rotary action again. 😵💫
Or I would just use the open ring configuration and not have to deal with it...
Well, I found the split ring thingie, now I need to add gear teeths to it.
... Or I could just put a little block/limiter on the parts of the body...
Sometimes I think too complicated solutions to problems that can be so simple to solve...
I also need to figure out a way of making the exoskeleton modular/adjustable on the linear parts for every body type...
Well, I did find one exoskeleton that have all the correct degrees of freedom of the shoulder and it doesn't have all the fancy thingies I previously suggested.
Although it is a medical exoskeleton, not a "power armor" type, it doesn't even have torso movement.
The only "problem" I have with that is that it has around 5 motors on the shoulders with the same amount of energy doing different things (I think).
You could have reduce that if you used the rotary stewart platform configuration where all actuators are working together.
Yes, it was just too bulky for a mech, but I think it could be a different story for an exoskeleton.
Potency and Power of Electric Motors:
Anyway, assuming that we need 300kgfcm (or 30 newton meters) of torque on each joint in order to lift 100kg (the weight of the user/pilot) and 30 rpm of speed, we would need around 94 watts of potency.
Triple that so it can lift 300 kg, and thus, need three times its torque, around 100 newton meters (or 1000kgfcm), so, 300 watts of power, around half a horsepower per electric motor.
On the shoulders and/or the hips you would have 6 electric motors working together, and thus, needing at best, 50 watts per motor.
You can find these 50 watts electric motors in the cheap.
Even if I used 400 watts electric motors, I would still use less than 1000 bucks for the legs alone (if I didn't use the rotary stewart platform).
The best part is that the electric motors I would need in this exoskeleton can be drone electric motors, these are super small and light, but have more rpm than torque, which would need a reduction ratio by either timing belts or reduction gears.
Well, taking into consideration the dimensions and capacities of this specific motor (link here), I would have 12210 RPM and 0.0782 Nm of torque at 100 watts.
And thus, I would need around 20:1 or 25:1 reduction gear ratio in order to achieve around 592.5 RPM at the end of said reduction in order to rotate around the rotary hexapod I posted in previous project logs.
Of course, these numbers will change accordingly to your setup and so on.
I think I will be using this 3D print file of a reduction gear, I can just copy the big gears and the smaller gears in a 2D print, cut it out of a medium and then glue both together.
Sigh...
This looks so stupid...
Why do I even keep insisting on this stupid idea?
It is so beyond my skills...
Well, I did find the R-3000 Rotopod 3D model online here, but it is paid; it wouldn't be a problem if I didn't want this project to be open source.
Edit¹:
But I was wondering if I should instead do "stewart legs" instead of correctly anatomical exoskeleton.
An example:
(there are better videos showing the WL-16)
Basically, it uses stewart platforms as feet-legs-knees-hips.
It is not as cool (and awesome) as walking exoskeletons, but it does the job (I think).
I say "I think" because I really can't see this thing stepping upwards...
Stewart Platforms are really versatile, but these don't have that much maneuverability (as a leg).
Edit²:
I think I will use this kind of actuator on the shoulder/hips because of the limitations of the stewart platform, specially because there isn't much space on the human chest for a long linear actuator.
Yes, these have a lot of friction, but I don't know any other mechanism.
(there is a link in the description with the files of said actuators).
Actually, I found a version with gears:
Also, I don't really need the screw drive, I can just use the rack and pinion on its place.
Sorry for the delay, I was trying to make something these last days, but for some reason, I just can't.
I just stare at the 3d modelling program and I just can't make anything out of it... :/
I'm still trying right now tho...
Good morning.
Right now I should be be preparing myself to do something, but I'm procrastinating.
Anyway, while I was trying different types of wattages and motors on those "Torque to HP converter" and "Pulley Ratio calculators", I noticed that I would just need an 900 RPM electric winch actuator since I would treat the limb itself as a pulley; and thus, the distance in which the cables are attached to the limbs accordingly to its axis of rotation will be more or less the diameter of a imaginary giant pulley.
(I need to remember to make an illustration showing that)
I hope the illustration is clear enough.
Also, I can't believe this, but I'm actually considering using the goddang micro servo motor "muscle" I suggested on Project Log 1.
These can use around 10 watts of power, but depending on its RPM, I would actually need half of it per motor. Some websites sell these motors with a rotation speed of 60 RPM, others as a 300 rpm. Dunno which is true, but hey... As long it is cheap...
I literally made more than 40 project logs just to go full circle and reach this stupid conclusion...
I hate myself...
Of course, before I definitely go through the winch path, I will try just one more time.
Because I feel like I didn't try enough with this rotary hexapod.
I found this "rotapod" from this website where multiple disks are stacked together.
I tried this configuration at the first time, but it ended too big for my taste (bigger than the previous attempts). But I will try with a more moderate sized parts and see where it takes me.
Too long...
Did it more like the first image. Kinda bulky tho...
Yeah... Too bulky...
Winch it is...
Of course, I won't be doing the exactly same idea I had before with micro servo motors, simply because those aren't fast enough.
Well, 5 of may will be my birthday, I hope someone suddenly appears out of nowhere with a free complete 3D CAD model of a mech... (as if).
Well, I don't really need to 3D model every gear set that I would need to use, but I do think it would be the best course of action taking into consideration my specific parameters.
For example, there are a lot of planetary STL 3D print reduction gears, but all of these use helical designs and don't take into consideration the low tolerances a 3D print has. So, basically, all of them can be really inefficient because 3D printers suck at precision printing, but none of them are concerned with that, because they simply make simple and small DIY projects.
So, I will need to find spur gears in various sizes and shapes so I can both use it on a 3D model and still being able to build it in real life.
Funniest part is that there are actually a lot of planetary spur gear reduction box STL files around the web...
Like this 40:1 compact gearbox.
I think I would only need to make the gears longer in order to be able to take more weight, or simply scale the entire model on blender...
There is also this one, although I don't know what is the reduction ratio (it says it is 6:1), it is a very solid design.
Well, I put the tooth numbers of these planetary gearboxes and I got maximum 4:1 in the first planetary gear, not 40:1, and 6:1 on the second.
The only planetary gear boxes that I can find that reach 100:1 gear reduction are like this, using helical gears and complex structures that I can't possible replicate with a simple setup like mine.
Well, since the last gear is a 35cm wide sprocket gear has 48 teeth, and the stages have a reduction ratio of 6:1, I divided the rpm value (36392 rpm) by 6 two times, until it achieved 168.481481481, then a 5 teeth round gear (like the used in the video of the rotary hexapod) that then it would achieve around 17.55 rpm.
Although the last stage kinda makes me confused... I mean, if you drive a bigger gear with a smaller gear, you get gear reduction, but since the smaller gear is the part that rotates, like in the hexapod case, doe sit get the same reduction gear? 😵💫
I will assume it does.
It looks a little bit better, kinda chonky tho...
Well, I tried to put it in a somewhat humanoid manner without everything crash on one another...
It seems like the hexapod idea won't work...
Winch/tendon mechanism it is...
Of course, I tried other way around, but it is just... Too large, I believe.
Okay, it is time to actually design stuff and 3D model my way out of this crap!
Okay, all right, I'm doing it!
C'mon brain, just work!
Another day passes and I'm still Mechless, Exoskeletonless, Exosuitless....
While I was procrastinating and having existential crisis, I started researching the power storage and power transmission systems out there.
And...
This container looking thing is a 1 megawatt battery pack.
It has 1 million watts and it would be able to power this mechsuit of mine for... 50 hours. Basically, two days of continuous use.
The bellow image is a 20kw electric generator.
Basically, a generator that is capable of powering my mechsuit for whatever hours...
I don't know if my brain is making pranks with me or anything like that, but...
I'm seriously trying to fit such thing in a backpack-sized power supply?
Me??!!
Remember all that talk about the 10horsepower micro turboshaft engine in previous Project Logs?
Welp...
There are inneficiencies on the generation of power by alternators and generators alike.
And for every 1 horsepower generated, I need to input around 2 horsepower to each car alternator... At best, 1.5 horsepower.
In the best case scenario, I would need a 37 horsepower engine to power 25 horsepower.
So, 4 micro turbines just to make the mech/exosuit walk.
Of course, I could shift towards porpuse-built efficient generators instead of stacking up generic alternators I would be able to excavange from old cars...
... But these are heavy, bulky, are the reason the electric generators are so big and heavy, and still can only achieve around 80% of efficiency.
And are also really fricking expensive...
By the way, althogh the idea of stacking cheaper and smaller generators with lower wattage ratings is tempting, the fuel consumption will proportionally scale with the amount of generators, that's why it is better to have a big generator producing X amount of kilowatts than a lot of tiny generators producing some amount of watts.
Yes.
I know.
I said in the project itself that mechs aren't efficient, neither pratical, and much less cheap.
But still...
It is kinda of a mood breaker to build such thing for nothing.
I wasn't expecting such thing to be this expensive...
Why I'm like this?
I say I will do something that isn't that great, and still, I say I will do it, and when I finally need to actually "do it", I start complaining...
In this section I had an entire page worth of rambling talking about an incredible "discovery" I had.
Basically, if you convert ethanol to hydrogen and then convert hydrogen to electricity using a fuel cell, you would get 30,000 watts of electricity for every 12 liter of ethanol per hour (around 35 to 40 horsepower).
There is a process called "steam reforming" that basically can convert any kind of hydrocarbon fuel (gasoline, methane, methanol, diesel etc) into hydrogen.
And for some reason, it generates more eletricity using "ethanol>hydrogen>electricity" pipeline than an ethanol fuel cell directly making electricity from ethanol.
The "only" problem is that a "cheap" 6000 watt hydrogen fuel cell costs 30,000 DOLLARS, costing 150,000 dollars in total if you choose to reach 30,000 watts of electricity.
So, even if one were to use hydrogen fuel cells with steam reforming, it would still cost more than literally just making an stupid turbine engine setup.
By the way, the reason Hydrogen Fuel cells cost so much is because they need Platinum to work efficiently.
Well, I guess if someone found a way of making efficient hydrogen cells without platinum, or even just an efficient way of making hydrogen from water, then I guess it would be enough to "hold the strings" while nuclear fusion is still in its infancy.
I mean, if we found a net positive way of producing hydrogen... It would almost be like nuclear fusion.
"Net positive" = you use less energy to make hydrogen than this amount of hydrogen is capable of producing.
Normally one of those car HHO generators uses 400 watts to make 1000ml of hydrogen gas.
So, accordingly to ChatGPT and the questions I've made on the internet that no one answered or called me an idiot, basically, a rod with 60cm of length should have at least 3 to 5 inches of diameter in order to sustain 20 tons of tension/weight being applied to it laterally.
So, I will make everything with 8 inches (20 cm) of diameter/size in order to stay in safety parameters, of course, assuming I'm using conventional UHMWPE and not the composite I talked about on Projct Log 45, which would double the strengths of the UHMWPE.
I was testing here using simple shapes with the proportions I intend to use, the big headless human mesh has 2,55 meters of height, and the tiny human on the side is more or less my height.
Okay, maybe the idea of using the same structure for the legs and the arms wasn't that great.
Besides, now that I put everything together, it feels like 20cm of diameter for everything is a bit of a overkill...
Now I tried these parts with a more moderate sizes of the parts.
It looks as beautiful as my face, and my face looks as beautiful as my ass.
That is just the structure itself, without any kind of electric motor, chain drive, reduction boxes, gas springs and counterbalancing springs.
Crap, I really took a project too big for me...
By the way, for the sprocket stuff I'm taking free STL models available on the internet, in this link.
In fact, I'm taking a lot of STL files because I'm bad at 3D modelling. :/
So, assuming that I will be using that 2300 watts electric motor for everything, this will be the gear box that looks like a servo motor box.
The nominal working power of the motor is said to be 1800 watts, but I will be using 750 watts for every one of the 6 motors on the hip "WHATS" joint.
And thus, the reduction will be from a 20mm diameter timing wheel rotating at 36392 RPM and a torque of 0.1968 Newton meter of torque to a 100mm (10cm) timing wheel that will rotate at 7278 RPM and 0.984 Newton Meter. Basically the top timing wheel in the image connected to the motor at the back.
Unfortunately, there isn't any kind of transmission system (such as gears) that I'm capable of building/3D printing, so it will be 4 timing wheels to support the 4 newton meters of power being transmitted for the next timing belt. Of course, since I will be changing the continuous force and speed every now and then, this system won't be able to handle much.
The smaller 4 timing wheels with also 20mm of diameter will drive the four bigger timing wheels with 10cm of diameter at 1455.6 RPM and 4.92 Newton meter of torque.
Then it wil drive a double sprocket/drive chain with also 20mm of diameter at 1455.6 RPM and 4.92 Nm to a 10cm diameter double sprocket that will rotate at 291.1 RPM and 24.6 Nm of torque to a last 20mm diameter sprocket.
This 20mm diameter sprocket will rotate the "WHATS" joint with 35cm of diameter with 15 RPM and 477.4 Nm of torque.
Well, this looks a little better than I initially imagined, to be honest... But still looks like my face.
Well, this thing is simply too bulky, I think I will have to mess with gears after all.
Actually, I believe there is a tool for gears on Blender, but I will have to look it up later.
I was interested on the project after the last Project Log, but right now my brain is saying that I should give up on the project no matter what. :/
... Or I could "just" put a 100:1 planetary gear on the motor and connect it to a 14cm radius sprocket and then to the 35cm radius sprocket on the base.
I would need to buy an original version and then copy it...
Now that I stoped to think about it, I think it could be better for the project if I treated the actuators like muslces. Yes, like the winch/tendon system I talked about in Project Log 1 so long ago, but it will only depend on my skill of making the gears.
At least for the legs, that will take so much weight that common gears/sprockets may not be the best choice.
The only reason I was avoiding using winchs was... That I completly forgot about them. :|
... And also because I wanted the hip joints work as a single actuator in all directions/degrees of freedom, this way I would spare a lot of electricity and a lot of money on the final system since I would different motors with the same power output to do different axis of rotation...
And on top of all of this, winch mechanism are incredibly inneficient. Every time I searched on google, asked on forums/websites or even to ChatGPT, winch mechanisms are bellow 40% efficiency (for some reason).
But if I follow the lead of a rope sprocket, I think I will be able to circumvent this dang efficiency problem.
Well, I made up my mind: no unproven designs like the electromagnets made out of electrical transformers sheets.
Electronic protectors have one of these inside themselves, and I found a couple of broken ones in my house.
I took like, two hours and almost bent all of them by simply taking then apart. These are so thin that sometimes they bent by simply falling on the ground.
Still usable tho, but I don't trust myself to build an entire electric motor with high efficiency and power.
I mean... I think I could build one, but a good one? I don't think so.
Counterbalancing Springs:
By the way, I just now remembered about counterbalancing springs on industrial robots.
Basically, a counterbalancing spring is a mechanism that supports the weight of the entire robot, so it only has to produce torque for lifting the payload and fighting inertia on abrupt motions.
So, basically, if the robot was to be slower, it would only consume half as much power, since half the power and torque requirement is to lift its own 1000kg weight, then... It one could cut its energy consumption and torque requirement by half.
... But I doubt it, it easier to counterweight a light robotic arm in a solid and stable base.
(I didn't finished watching by the way)
...I think I finally understood how these springs work (I finished watching it).
The springs are meant to only counterbalance the weight of the robotic arm itself on its highest load capacity instead of any situation no matter the position.
So... It is not ideal for other uses, like a walking robot. It is only capable to sustain its own weight, not to dynamically change its counterbalancing on real time.
Of course, it is not that simple.
How to counterbalance an entire fricking humanoid mechanism?
If it was, then we would already be seeing passive exoskeletons where the entire weight of the body is absolutely supported by the exosuit.
Which is not the case.
So, yeah, what I previously wrote must be incorrect or simply wishful thinking.
I just now remembered the "leg on air" problem (it is not a real name btw, lol).
Basically, if you counterbalance a single leg of the robot for the weight of the entire body, what will happen when this counterbalanced leg lifts off the ground and doesn't has any loads on it?
The 2 tons counterweight will simply go "boing-oin-oing" and pull the entire leg to its direction.
Not to mention that once you divide the weight of the entire body between two legs, the counterbalancing spring will be weaker when the entire weight of the body is under only one leg...
One would need a real-time counterbalancing system that changes its load capacity.
And the only system that I can think of is a pneumatic system, because I really can't see a spring changing its load unless it is mechanically.
A good way to start looking into this system would be quasi-passive exoskeletons.
It was said in the article (I looked at it using Sci-hub) that they achieved a 30-40% diminishment in the effort required to walk.
Well, not 50%, but a cool amount nevertheless.
However, as you can see by the image itself, it doesn't look like a simple work to achieve...
I was thinking on having a tank with pneumatic or hydraulic accumulator linked to each gas spring (or individually for each gas spring) using a spring as the compressor/pressurizer in the accumulator, with everything with the equal pressure.
So, when the limb needs to move, a solenoid valve opens up and allows air to pass through to the accumulator, when the limb is going to stay still or absorb an impact, the solenoid closes up.
The air will move in and out of said accumulator due to the effort of the electric motors, if it is going to help or stop the electric motor will be a matter of controlling the opening and closening.
Of course, you would need to program it, and I would do it in the simplest way possible (like, "open valve X when electric motor Y moves clockwise"), so I doubt I would be able to achieve 40% less load on each limb.
Besides, like I said before, it will require extra energy for abruptely change the direction of said limb. But... I will try to make it as light as I possibly can, using even densified wood and epoxy.
Yes, you could add compression springs/hydraulic dampers on the feet of the mech and so on in order for it to absorb the change in direction and then, finally adding the force of the actuators.
However, it would still be interesting to have the last resort of using some extra torque from the electric motors to avoid the worst scenario possible. You could also overload your actuators and pass more energy than it is supposed to receive at the risk of damaging it, but I digress.
Although a change from 6kw (8hp) electric motors to 3kw (4hp) electric motors would be welcoming to my wallet, lol.
Of course, I would need to go designing limb by limb, screw by screw.
And that's why I made this Project Log. I need to start first and then go from there.
So, like I said in previous Project Logs, I will try to use the same mechanism in all parts of the body of the mech, so it is easier to build it and build extra parts. Because, in the end, if you use the same electric motor, joints and limbs on everything, you won't need to build costumized/unique parts.
So... Why the "WASH joint" thingie?
Because a Wrist is a joint with 3 degrees of freedom, an Ankle is also a DOF joint (ok, not the human ankle, it can only rotate on 2 axis and barely on a third one [unless my leg is messed up]), the Torso and the Shoulder is actually a 5 degree of freedom joint (because in a 6DoF parallel mechanism, like the Stewart platform, you can go up and down on its own axis, unlike our shoulder), so it is basically two wrist/ankle/hip joints facing eachother.
Here a visual explanation of what a "degree of freedom" is. When someone says "X number degrees of freedom", it is a robot that has a movement on one or more axis shown above. You could make a 49 DoF robot that only moves in the horizontal plane, for example.
This means that by making a 3DoF hip-ankle joint that can withstand the entire weight o the body without falling apart, then I can use the same exactly mechanism on other parts/joints of the body.
At first I was thinking of making this kind of mechanism, that is a rotary parallel platform (I think this is how it is called), but as you can guess, it is too complex to acurately be planned by an amauter like me.
There is a 3D printed version (I hope the entire playlist appears once you click on the video.
But what made me interested on the design is the fact that all three electric motors would be working at same time to perform the same action, thus reducing the load on each actuator.
But I really don't know how to make such thing to support 2 tons of weight and 20 tons of torque...
There is also a 2DoF robot eye, but it is not rotary. However, if I was a good engineer (which is not the case) it could be worth the trouble making a 3DoF robot eye with just 2 actuators.
Maybe if you remove one of the axis crom the previous example and adjusted its angle so it could only have 2 actuators... Maybe.
And funny enough, there is already an easily accessible version that is already commonly used.
And the last one is showing being used as a hip joint.
And yes, I intend on using either belt drives or cable/pulley drives on everything (because it is cheap).
And yes², this also mean that I won't need to design as much I expected myself, because the biggest part of the design (the WHATS joint) is already available in 3D models or STL models for 3D printing.
The only down side is what I said previously, that the parallel manipulator platforms would have all 3 actuators working at the same time, unlike this new one.
I "just" need to find out how much material each limb will require in order to sustain the 20 tons of torque, design the spring counterbalancing system and the "soft locking" hydraulic or ratchet system..
Now... Time to design.
... But before that, check out this incredible belt drive reducer:
I didn't quite understood how it works, but this is really creative and fascinating.
It is like a planetary gearbox for a belt drive. :O
Ah, here is his video explaining it, lol
(I still don't get it, but ok)
(I really don't get why there must be a pulley on the bottom that doesn't move)
Actually I didn't think orbital gear belt drive was possible, but it actually is:
Torque and other mechanism limitations:
I planned on deleting the videos above because of this section, but for some fricky reason, Hackaday website just bugs out and don't allow me to delete these.
Well, anyway...
While I was looking at electric motors and reduction/transmission mechanisms, I found out a couple of things:
I wouldn't ctually need 8hp/6kw of power in the joints/limbs of the mech/exosuit, but around 5hp/3750 watts, the difference is the amount of RPM.
But it will depend heavily on the design of the limb/actuator.
For example: If the exosuit is for my size, and its hip joints is 90cm distance from the ground, then it would need to have a different torque from a mechsuit with a hip joint 1,20 meters (120cm) away from the ground.
Of course, if you are using a cable/winch/tendon mechanism, then you would need a different amount of torque.
Soooo... I keep looking/testing different torques and distances for each size, and I feel like I'm probably exagerating a lot on the amount of torque required.
To be honest, it has been so much time I forgor why the heck I need 18 tons of torque. :|
I remembered why, because if I want to lift 2 tons (the weight of the mech itself and its load), I would need to lift at least 3 times as much, and thus, 6 tons. And since each actuator is in an disadvantage of at least 1/3 of the length of each limb, I would need to lift 3 times as much, and thus, 18 tons.
I was looking for humanoid robots such as Atlas from Boston Dynamics, and it is said that each one of his leg joints can achieve a torque of 150 newton meters, which is more or less 1500 kgfcm of torque.
So, if you divide the amount of times 2 tons is bigger than 80kg (the weight of Atlas), then you would see that this mech would need 25 times the amount of torque of Atlas; and you would get around 3750 newton meters of torque, which is almost 40,000kgfcm.
And thus, one could say that this mech is under powered compared to Atlas at its full weight + its load capacity (1 ton).
The 20,000kgfcm of torque is capable of lifting its own weight, but not its weight under load (supposedly).
The other thing is that I found out that some mechanisms have a limit of speed and torque limit they can handle/transmit.
For example, a belt drive has a limit of 50 m/s of speed and a torque limit of 1 newton meter of torque (around 10kgfcm), depending on its type and so on.
But this is a really generic amount of speed and torque, and the previous videos show different amounts of torque and speeds. For example, the first video with the "NEMA17"belt drive can only take 500 grams of load before skipping teeths, which would be 1.4 newton meters (it is a direct connection from belt drive to the arm, so I don't know how valid is my concern)... Plus, I will only have access to also generic transmission mechanisms.
Better safe than sorry, I guess...
So, if I were to use those 2400 watts electric drone motors, I would need to carefully change its timing belt reducers in order to not exceed these speeds. And even change from timing belts to sprocket chains.
The best way I could find to check if I didn't exceed the speed limit of belt drives, I just enter the diameter of the toothed pulley on a tip speed online calculator and it shows how many m/s it achieves.
On top of that, I kinda check on this other online pulley calculator to see the ratio/diameter/torque of each stage.
And I would need around 9mm of diameter toothed pulley on the first stage to reduce 100,000 rpm to around 20,000 rpm in a 4cm diameter pulley to stay around 47m/s.
However, these are just estimatives, and I'm barely at the limit of each material, and I don't like to be at the limit, because the wear and tear of everything will be a bigger problem.
Of course, I could add more belts to each stage to better distribute the 1 newton meter of torque limit and so on...
But it will depend heavily on the final design and the final torque/rpm required.
For example, I could treat the entire limb like a final stage for the pulleys, and thus reduce 1 to 2 stages depending on the motor (I would need around 4 to 5 stages in total).
And since the "WHATS" joint has 3 motors working at same time for every movement, I would actually need 5 horsepower (3750 watts) in total, dividing the load between the 3 actuators, and thus, I would need less expensive/powerful/energy motors to something around 1250 watts.
Of course, if you have a 6DoF WHATS joint with 6 electric motors each, then you would need even less energy to move that part of the body. Around 625 watts per motor, and a horsepower is 750 watts.
Manufacturing Process:
Dunno if I should be talking about materials this early, but I was thinking on using resin with densified wood or UHMWPE (Ultra High Molecular Weight Polyethylene) with densified wood and/or nylon.
Simply put, UHMWPE has a good tensile, compressive and impact strength for its price, density and easiness of work. After all, how do you melt/join two pieces of steel? And how you would to the same thing to UHMWPE? You can melt polyethylene with a hair dryer, steel, in the other hand...
Plus, if you use laminates of UHMWPE and melt-press them with something as simple as a clothes iron, you can make it a little bit more resilient.
Specially if you throw something to occupy the plastic matrix, such as wood and/or other types of polymer, such as Nylon.
(just now it came to my mind that wood won't be able to withstand the temperatures to melt UHMWPE)
And yes, there are composites of both materials. Some are even bulletproof.
Also, there are something called "Duroplast" used in cars and on toilet seats, which is a resin mixed with cotton waist and stuff like that.
But I couldn't find the actual "recipe" for making it, neither how actually strong it was.
Welp, I can just throw everything in the "caldron" and hope for the best, lol.
Obviously, I won't literally "hope for the best", I will try to study/verify everything as humanly possible to my habilities and then make this crap happen.
For example, using the information available on the types of polyethylene, like its young's modulus (its rigidity), its tensile strenght, compressive resistance and its impact resistance.
For example, this articles shows a pure composite material made using UHMWPE and PE wax that greatly increases all these properties.
I know that earlier in this same Project Log I talked about how I wasn't confident on making stuff by myself.
... But just now I remembered that there is a technique involving normal ink jet printing and laser cutting.
Basically, "just" print a 2D image of the piece you wanna make (like a gear, an artificial bone etc) on top of a block of what you want to make and "just" cut the excess.
Well, not to mention that since I will be literally using plastic as the building material, this means I can hire laser engraver/cutting stores to cut the solid blocks for me in the shapes that I need.
Also I was thinking on DIYing a mix of drive-chain with pulley, so to speak. Basically using UHMWPE rope (that is cheap and can sustain 6 tons per meter) to make "holes" in which the teeth on the sprockets would pull them.
My reason would be using the strength of UHMWPE in the tension strength department as my advantage on top of allowing some kind of non-desctructive gear slippage (I think).
... However, I couldn't even find a name to this kind of drive, the closest I could find was "perforated belt drive", which Google only shows[ belt drives that work normally and have holes on it for other reasons than to be driven by the holes.
The closest would be the rope ladder on the right.
Design:
Anyway, DESIGNING:
(sorry if I cut the explanation short, but I feel like I put too much stuff in this project already)
Well, first:
Although I said "I wouldn't have to redesign much", I actually need to redesign something.
As you can imagine, not every bearing you can buy online or build at home will be able to withstand thousands of kilograms. And as such, I can't really trust on the structural integrity of everything I can build.
I mean, just watch the start of the Capstain Drive up there (at 1:52 more specifically), just 3kg weight at the tip of the arm was enough to bend the connection between the drive and the metal beam.
I just now remembered that I'm bad at 3D modelling...
Like I said in previous Project Logs, I calculated that I would need around 6000 watts electric motors (or 8 horsepower electric motors) in order to be able to lift 1000kg of weight and 1000kg of the mech's own weight.
The problem is the same as always: money.
I simply can't find 6kw electric motors, and I can't even find any kind of electric motor that was specifically made for low rpm high torque in the area of 6kw. Only those gigantic ones that weights 2 tons.
And the problem is that I can't just take 6kw rated electric motors, just like combustion engines, I have to take motors that are a little more powerful and then use use with 6kw of power.
Simply because if I take a 6kw motor and keep using it at 6kw, it will put the motor at its limit and overheat, losing efficiency. Every motor has a nominal working load.
But I could only find 15kw motors or above, and these cost around 4000 reais (800 dollars).
Well, this time I think it would be relevant to specifically state the name, power ratting, price and image of every motor I come accross.
This one is called "Turnigy RotoMax" and it comes in a variety of power ratings and all of the ratings are expensive as hell.
Some are just 2000 watts of power and cost thousands of brazilian bucks, some are actually 10kw or above and also costs thousands of brazilian bucks.
But depending on the place, I can find its price ranging from 2000 to 5000 reais (200 to 1000 dollars).
(funny enough, I found one with 25kw rating costing 450 dollars or 2200 reais).
While searching for high power E-motors, I come accross motor-wheels meant for electric scooters or things like that.
Those have some serious power and some serious costs.
This one in specific is 15kw of power and also costs 4k reais (800 dollars) and accordingly to the aliexpress ad, it achieves 146 kph in a e-scooter.
Well, there are a lot of other types of e-motors that I come across, and all of them look exactly the same and costs around the same amount.
I need at least 3 to 5 electric motors for each leg, so, if an actuator costs 2000 bucks, I would need 20 thousand reais to 50 thousand reais (4k dollars to 10k dollars).
Not to mention that I kinda have the intention of making "repeating" mechanism all around the mech/exosuit, so I just have to build the same thing for the arms, for the legs and for the torso.
But this would mean buying the same motor for every part of the body and jut changing its energy supply.
... Which would cost... A lot more...
So I thought on buying those smaller drone electric motors that have around 2400watts of power and "simply" stack these together in order to achieve 6000 watts.
There are a lot of different types, sizes and costs for drone motors, so I will just talk about the cheapest I found for the nicest amount of power, so to speak.
This one is called "Turnigy XK3674" and I found its ad on the website "hobbyking", it costs 189 reais (38 dollars).
This one in specific has a power rating of 2400 watts, 70 amperes, 33.6 volts (which, if multiplied by each other to find the wattage, it is actually 2300 watts) and 1650 KV (which is like 115500 rpm).
I would need 3 of these for the 6000 watts of power and it would cost 567 reais (115,47 dollars) in total for each group of actuators. Costing in total 5670 reais (1150 doolars) for all the 10 actuators on the legs...
But...
However, a subject that I was avoiding until now was precisely the reduction gearboxes that one would need to extract 18-20 rpm out of these actuators.
Using an online reduction ratio calculator, I found out that I would need around 4000:1 reduction ratio from 115k rpm to 18rpm (it actually is like, 3993:1 or something like that, just rouding the numbers).
If you put the bigger number on the input on the calculator and then a smaller number in the output, it will give a number like 0.0005:1 ratio, but since the sellers use the "100:1" ratio, you need to insert the numbers in reverse on the caculator.
And one interesting thing about gear ratios is that if you take a 100:1 gearbox output and insert on the 100:1 input, the output will be multiplied by the number of ratio reduction, so, 100x100 = 10,000 ratio.
So, in order to achieve a 4000:1 gear ratio you could take a lot of different gears ratios in order to achieve the same gear output, so, a gearbox with 40:1 gear ratio and then connect it to a 100:1 gear ratio = 4000. A 60:1 gearbox to a 60:1 gearbox, you will get around 3600:1 gear ratio.
Unfortunately, these don't really have a name, but the biggest and the most compact gear reduction ratios are achieve by planetary gearboxes meant for stepper motors.
And these are more expensive than the electric motor itself.
One of these cost 600 reais (122 dollars) and I would need 2 of these per motor...
The 3 electric motors just costs 567 reais...
There are more expensive motors than these, but let's agree here that I really doubt this little thingie would be able to have a 20,000 kgfcm and 18 rpm.
I also had the "bright" idea of using actual combustion engines attached to smaller reduction gear ratios with reverse gear, so I could make the limbs go back and forth.
These cost around the same price as the 15-25kw electric motors, and weights around 20 to 40kg.
But I quickly abandoned the idea, I don't even know how to manage the electric starting of the engines and so on...
So... What this brings us to?
Well... I had the "bright" idea of taking old/broken transformers and use the silicon steel of these things and make an electromagnet.
Transformers already are electromagnets, closed loop electromagnets in specific, so one would "just" need to open it up.
However, these have isolating laminates between the metal laminates. So I would need to go one by one taking these out.
At least the laminates are so thin that I think I can cut these with just a scissor.
Also, conventional electric motors already have laminates, not in the same level of a transformer, but they have.
The principle of laminating metal cores of electromagnets in electric motors is exactly the fact that they turn on and off really fast, which can leftover electric currents (eddy currents) on the core, which can be transformed into heat.
This guy makes an electric generator using old transformers, of course, a generator is just a reverse motor.
By the way, this is just one of the thousands of videos on youtube that say things like "DIY 5000 watts electric motor/generator", never shows the current and the voltage and makes a sketchy looking thing that probably wouldn't be able to even output 50 watts.
I would need to copy existing electric motors the best I can to achieve this 6kw electric motor.
Then I would need to find out what is the best AWG/wire diameter, and then find out the best structure for the neodymium magnets or simply an electromagnet.
I would also need to find the ideal halbach array that allows for the best electric rotor.
(the B type is ideal for cars, or so I heard)
And on top of all of this trouble, I need to find a way of making a generator/alternator that is compact enough.
All 40hp (24000 watts) generators are simply ginormous, but taking alternators from cars is also an alternative.
This guy stacked a bunch of alternators in order to achieve 30kw of power...
... For powering speakers...
There are some DIY videos on the subject.
Although, with such low rpms of 18 to 30 rotations per minute, I would guess that the "best" option for me would be stepper motors, but I don't think I will be able to build these. >.>... And after all of that, I don't even know if it would be more cost-effective to simply buy the goddamn 2k electric motor. :/It is not like there are online tutorials on how to make 30 rpm 20000kgfcm of torque electric motors out there tho...
Meaning that I just need half of the wattage power I calculate for each electric motor (and once standing still, or moving both legs at same time in a jump, the torque/weight would be divided for each leg).
Meaning that I don't need 420-500 horsepower, but actually "just" 200.
And in the new calculations I've made here, I would only need half of its power (25, 105 and 40 horsepower).
I'm angy, but also happy, because, although this project was lost, it actually is fine. :)
So... I was verifying some stuff about the electric motors and the electromagnets and surprisingly, I think I've made some progress.
Electromagnets:
Some people on quora said that "a holding electromagnet would lose more or less half of its attraction power in a 1mm distance from the electromagnet assuming it is on the presence of air" (air increases the resistance to electromagnetism, somehow...?).
So, instead of the probable 20 horsepower, it would be the hyphotetical 40 horsepower.
Not bad, not great, but better than 400-500 horsepower.
Also, I was wondering if connecting all the electromagnets to the same cable/rope would be of any use.
Like in the sketchy illustration above, where a cable goes in a straight line and then connects to the electromagnets.
But the most probably scenario is that the top and bottom electromagnets would separate due to its limitation (since these can only hold 150kg instead of 600kg) and then the electromagnets would progressively separate once its limiting holding strength is individually surpassed. Like a cascade event.
But I don't really know if it would be 100% accurate since all of the electromagnets would be holding each other at once.
But I can't even test this crap out at my home because I'm too broke to buy electromagnets (even the cheaper ones) or to buy one of those DC powering stations in order to properly power homemade electromagnets.
These cost around 300-600 reais (60 to 120 dollars).
You may be wondering: "if you're too broke even to buy a small setup, what makes you think you will be able to build an entire mech?".
And the answer is:
I refuse to think about the consequences of my life choices.
In any case, I would suppose that this electromagnet artificial muscle is, unfortunately, a dead end. :/
Edit⁴:
So... I was researching about the reversely actuated elastic elactromagnetic muscle.
Basically, it is the same thing as this artificial muscle, but instead of using pneumatics/hydraulics, it uses electromagnetism.
So, the idea is that the air inflates, and thus, extends the elastic rubber material. Once the air is released, it quickly goes back to its original state. The idea of an electromagnetic reversely actuated artificial muscle is the same as this one, but using electromagnets.
Basically, I would have sheets of latex covered with electromagnets, while the electromagnets are attracting each other, these would squeeze the material. Using less energy than the pneumatic version (supposedly).
So, I asked around trying to figure out how many latex sheets/bands/tubes I would need in order to lift 1000kg or 20000kg, like in the 1 ton lifting capacity mech.
(Actually, the maximum length is 119mm and the diameter is 16mm, I calculated the inner diameter alone)
If one were to convert the circumference of the diameter to width, then this latex sheet would be 3cm long and 4cm wide.
If I tried to simply multiply the values until it reached the sheet dimensions that I mentioned in the question, it would be 10 times longer and 5 times wider.
If I were to directly multiply these two by the 8kg lifting capacity of the latex material, I would achieve a 400kg capacity of each latex sheet. Meaning that I would only need 2.5 sheets (let's round up to 4) of elastic latex with 30cm of length, 20cm of width and 1.6mm of thickness.
If I were to convert the dimensions of the sheet with 30cm of length and 20cm of width, it would have a surface area (of one side) of 600cm².
Since the muscle applies a 58 PSI or 4.07 kg square cm of pressure on the surface of the inner tube, I would assume that in order to squeeze the latex sheet with 1.6mm of thickness using electromagnets, I would need to apply a pressure of 2442 kg on its entire surface in order to make it extend almost 5 times. But I don't need 150cm.
I just multiplied 4kg force applied per square cm of area (or 4 kgfcm²) and multiplied by 600, I hope it is correct.
A 3000kg holding electromagnet uses 300watts (accordingly to a random ebay page). And since said electromagnet has a diameter of 240mm, it has an area of 452 cubic cm. Which would more or less cover the 600cm almost completly.
Assuming that this electromagnet squeezes the latex sheet well enough, and since I would need 80 of these in order to lift 20 tons, I would need 24kw of power (32hp) to squeeze a single reversely actuated electromagnetic artificial muscle.
Even if I cut the required power in half because I don't need almost 5 times its length, it would still be 16hp, and even if I cut it in half again because of the thickness of other factors, I would still need 8 horsepower per muscle.
Not much different from the other electric motor alternatives.
All of this in rough estimates, of course.
I would guess that electromagnetic artificial muscles is super dead to this project (if I'm correct).
Of course, I can be really wrong, or really right, even if this version is simpler to build, it is not a trustworthy option/not a proven design, unlike electric motors.
Sooooo... I gave it a try and I tested linear and solenoid electromagnetic actuators. But unfortunately, these would consume the same amount of power as an electric motor.
The idea was to move the solenoid coil itself through a soft steel core rod instead of the rod.
My thought proccess is that the soft steel rod would be heavier than the coil itself, so less mass to move. And on top of that, normally the shafts tend to move to the center of the solenoid coil. So I thought that it would be simpler to actuate/program a solenoid coil since its position would always be the center of the shaft.
But I don't know much about the subject to afirm such thing with propriety.
Solenoids need springs in order to return the actuator to its original position, in this case, off-center.
Well... In case I actually build a 500 horsepower turbine engine, I suppose this would be the easiest approach for me. Since I have no fricking idea how to make a 48hp electric motor that has 180rpm and 18000kgfcm of torque.
RPM x Actual Speed:
So, something that actually kinda gave me hope for the solenoid and the electric motor was the actual speed of the object.
You see, RPM is the amount of rotations that a given center of a shaft rotates per minute.
However...
If you stop to think about it, the rotation of the center of the shaft is not the speed of rotation of the border of the disk.
If you were to rotate a 20 kilometer diameter wide disk at mere 1 rpm, its tip would be breaking the speed of sound.
The circunference of a 20cm diameter wide circle is around 62cm long, and if you remember what I said in previous Project Logs, if you take the radius of a big wheel and a smaller wheel, it would be like building a mechanical lever.
Summarizing: a gear/wheel is just a rotating lever.
This means that I don't need 180 rpm to achieve full actuation of arms and lengs.
For example, if I just change the angle of the limbs "wheel" from 180º degree angle (in neutral position) to 45º degree angle (completely actuated) I wouldn't need to fully rotate 62 cm, but around 1/4 to 1/6 of that.
I feel like I'm being unclear here, but let me try to explain:
See this rotary gear to linear gear.
If one would take a gear with 10cm of radius, and thus, 20cm of diameter and rotated a 1:1 ratio linear gear, it would travel 62cm distance per rotation.
And since it rotates 3 times per second (and thus, 1/3 of a second), it would travel in a straight line around 180cm.
... But I don't need 180cm of linear distance per second, I need to travel from 6cm-10cm to 30cm per second.
... And since I "just" need to change the rotary action to linear action (or the reverse), I don't actually need 180 rpm, I just need 18 to 6 RPM.
... Meaning that in order to achieve 18 tons to 20 tons of force I don't need 180 rpm, but 10 times less rpm than that.
... Meaning that I don't actually need 48 horsepower per limb, but only 5 horsepower or 3.6 kilowatts of power per limb.
Which would be 50 horsepower in total, or 36 thousand watts in total for the lower body. Aka, leggs.
And I think the logic is also valid for the human muscles, if you really contract your biceps and measure more or less the distance it takes from 180 degrees to touching your shoulders, the muscle itself just travels around 5cm (in rough estimates).
Other kind of rotational speed unit of measurement is radians per second.
If you change your limb position from 180º angle to 45º angle, you would be traveling 155 degrees.
1 RPM = 57.29 radians per second.
Since I'm traveling 155 degress in a third of a second, then I would be travelling 8.11577811986 radians per second.
Which would give 77.4 RPM.
Which if you put in a RPM Torque to horsepower calculator, then it would be 15 kilowatts of power, or 21 horsepower per limb.
Giving 210 horsepower in total for the legs.
Which is half of what I calculated in previous Project Logs, but I don't know which one is the right answer.
Edit¹:
I know this isn't a very super accurate measurement, but I tried to rotate my entire leg from 180º (standing) to 90º angle (on the horizontal) and back to standing as fast as I possibly could, to the point it kinda hurt, lol.
And basically, I was able to to do this in a more or less a second, which would be 180º angle of travel in total, which would be 3.14159153027 radians per second, or 0.5 Rotations per second, or 30 Rotations per Minute. Not 180 rpm.
Which would be 6.1 kilowatts of power, 8 horsepower per limb.
Or 80 horsepower in total.
Not the greatest result, but even in full power, it is still better than 400 horsepower.
I still don't know which result is the best/correct, but I'm sure more optimistic about this project than before.
This one will be a really short Project Log, simply because I want to "officialize" and announce to the followers (since I literally made another Project Log inside of the previous Project Log in the form of edits) that I found an interesting alternative to electric motors.
Edit²:
It seems that things will be okay-ish, I've made this illustration of the electromagnets in a "Z" or "staircase" connection configuration, and it gave an interesting result. It seems like it would make a twisting action on top of attracting each other:
Of course, I just twisted the bases because I thought it would be clearer to see the thing happening.
But I'm still not convinced, see:
When the squares representing the electromagnets connect to each other, it makes an "U" shape. And the cables that aren't touching or passing through the top/bottom plates would just be hanging loosely.
This means that I should have the bases in a "U" shape in order to make every electromagnet working together?
Or that I should put something flexible in the place of the plates, so it would actually work like an actuator?
Edit³:
I found this scientific article that explores an electromagnetic artificial muscle, but for the life of me, my attention deficit isn't allowing me to read the damn thing.
This means that I can simply attach every single electromagnet to a rope/cable and all will act at same time to pull the rope?
Edit¹:
I just remembered a fatal flaw with this idea of stacking up electromagnets:
Their force won't stack with it.
If you stack electromagnets one above the another, their force will stay the same, so their lifting force will be limited by its own strength.
Meaning that even if I stack them up, they will separate once the load goes above their limits.
I know I don't have a big brain, but goddang, sometimes I outdo myself.
Now I need to figure out a way of using electromagnets it said limitations.
The only three ways I could think of a solution was:
1- Using something similar to reversely actuated artificial muscles, let's say, the electromagnets squeezes rubber until it extends and then these disactivate gradualy in order to let the rubber actuate by itself.
2- Copy how the HASEL actuator works, but with electromagnets, the problem is that I don't fricking know how well it will work, since it would basically be an hydraulic muscle activated by an electromagnet. :|
3- Somehow, someway, distribute the force of all electromagnets throughout a line of metal plates parallel to the electromagnets.
Yes, this is an illustration of a dielectric elastomer, but imagine that one side of the "m" are the electromagnets moving 1mm each and the other side is a bunch of metal plates connected to a steel cable that will also move 1mm each.
On Project Log 40, I calculated a little bit how many watts of energy I would need in order to achieve the lifting capacity of 100kg and 1000kg, and in the case of the 1000kg, I found out that I would need 300 kilowatts of power, around 400-500 horsepower.
However, after looking at the electromagnets that are meant to lift 150kg and its power consumption of just 12 watts, I got the idea of just using a bunch of electromagnets stacked one over another in order to lift various tons.
Funny enough, I only checked this kind of electromagnet because I was looking for ways of making the electric motors cheaper.
Of course, it is not that simple.
If you simply stack one electromagnet over another, you will lose electromagnetic power for weird physic reasons.
"This electromagnet (two electromagnets one above the other) would be like one with twice as many turns of wire in its winding. Since the magnetic field produced by a solenoid (and an electromagnet is nothing more than a solenoid with a soft iron core) is proportional to the number of turns, it would produce a magnetic field twice as strong. In that case, however, this doubled number of turns would occur over a length that is also doubled, which would reduce the magnetic field by half. Thus being doubled by the number of turns and divided by two by the length, the magnetic field would remain the same. Being able, therefore, to exert only the same force, that is, in this case, to lift only ten kilograms. But if they were placed side by side and, no, one behind the other, then, yes, they would be able to lift twenty kilograms."
So, you would need to screw an electromagnet to a non-magnetic plate (like wood?) and then point it to a metalic plate that would also have a wood plate and an electromagnet and so on until you achieve whatever tons of power you need.
I also got confused if my logic was correct:
"If one were to stack 50 electromagnets one above the another in a 1mm distance, if all of these electromagnets were to close that distance to zero, would the entire stack of electromagnets shrink in 50mm or just 1mm?"
ChatGPT:
The entire stack of electromagnets would shrink by 1mm, not by 50mm. This is because when the distance between two objects decreases, the effect of that decrease is localized to the immediate vicinity of the objects and does not affect the entire length of the stack. Therefore, each electromagnet would move closer to its adjacent electromagnet by 1mm, resulting in a total stack length reduction of 1mm."
Then I tested the hyphotesis on Blender:
If you can't tell what is going on in this screenshot, let me explain:
The bottom cube has 1 unit of height, I applied an "object modifier" on the cube so it self replicates 50 times in a direction of my choosing.
I choose it to replicate 50 times in the vertical direction and then inserted a distance of 2 units from the center of the cube, when you change this distance from 2 units to 1, it changes its verical size to 50 times.
(by the way, I tried to set the point of multiplication for the surface/border of the cube, and the result was the same)
And thus, if you replace the distance from 1 unit to 1 milimiter... You get the idea.
I copied and pasted the same cube to the side so I could properly measure the distance between each cube.
So, yeah. This seems promising.
Now I just need to figure out what materials I would need and how to avoid the electromagnets from magnetically interferring with each other.
Also, In the case of the improvised electric motor with these lifting magnets, I was thinking of using them as a kind of stepper-motor.
Inverting the polarization of some magnets and keeping the polarity of other magnets in order to use the full potential of all magnets at the same time.
So, even if I had 60 electromagnets with 150kg of power working, I would have a holding torque of 10 tons using a fraction of the power of an electric motor.
But I don't know how well this idea would work since I don't really understand much about electric motors.
Dunno if this is a relevant edit/information, but I was looking around and it seems like maybe electromagnets may be the new hope for this project.
In summary, if I were to take those 150kg holding force electromagnets and stacked one above the another, I could lift heavier weights with less energy.
If we assume that one needs to lift 10 tons with these magnets, one would need around 66 of these.
Which, in total of 20 electromagnet muscles, it would consume around 16kw, with is around 21 horsepower.
Of course, it has its details.
The electromagnets are for holding force, not attraction force.
Although one could separate these in 1mm distances until the sum of all the spaces is around X centimeters.
If there are 50 1mm spaces between 50 electromagnets, then it would move around 5cm (50mm).
Edit⁴:
I asked ChatGPT:
"If one were to stack 50 electromagnets one above the another in a 1mm distance, if all of these electromagnets were to close that distance to zero, would the entire stack of electromagnets shrink in 50mm or just 1mm?"
ChatGPT: The entire stack of electromagnets would shrink by 1mm, not by 50mm. This is because when the distance between two objects decreases, the effect of that decrease is localized to the immediate vicinity of the objects and does not affect the entire length of the stack. Therefore, each electromagnet would move closer to its adjacent electromagnet by 1mm, resulting in a total stack length reduction of 1mm."
As you can guess, I'm not the brightest (assuming ChatGPT also didn't make a mistake).
So, I tried it on Blender, and it seems that ChatGPT was incorrect and I was right. dang.
If you can't tell what is going on in this screenshot, let me explain:
The bottom cube has 1 unit of height, I applied an "object modifier" on the cube so it self replicates 50 times in a direction of my choosing.
I choose it to replicate 50 times in the vertical direction and then inserted a distance of 2 units from the center of the cube, when you change this distance from 2 units to 1, it changes its verical size to 50 times. (by the way, I tried to set the point of multiplication for the surface/border of the cube, and the result was the same)
And thus, if you replace the distance from 1 unit to 1 milimiter... You get the idea.
I copied and pasted the same cube to the side so I could properly measure the distance between each cube.
Of course, you can't use any kind of metallic surface. You need ones that are good for electromagnetim. Also, you could use neodymium magnets to work together with the electromagnets (I think).
But I still don't know if stacking one above another would work at all.
Edit³:
Asnwer I got on Quora:
"This electromagnet (two electromagnets one above the other) would be like one with twice as many turns of wire in its winding. Since the magnetic field produced by a solenoid (and an electromagnet is nothing more than a solenoid with a soft iron core) is proportional to the number of turns, it would produce a magnetic field twice as strong. In that case, however, this doubled number of turns would occur over a length that is also doubled, which would reduce the magnetic field by half. Thus being doubled by the number of turns and divided by two by the length, the magnetic field would remain the same. Being able, therefore, to exert only the same force, that is, in this case, to lift only ten kilograms. But if they were placed side by side and, no, one behind the other, then, yes, they would be able to lift twenty kilograms."
So, I need to find a way of making all these electromagnets to work together without interfering with one another....
I do have time right now, but I'm really not in the mood to calculate the precise torque/energy requirements of the electric sub-motors of the exosuit/mech.
I tried a few times and these actually need more power than previously calculated, not less.
Of course, it is not that easy to calculate (at least for me, an illetrate on the subject) since you wouldn't be always needing the same extremely high torque and speed all the time.
For example, it is much easier to carry a load once the load on your arm passes over 90 degrees angle and goes to around your shoulder.
Translating that power requirement to a robot, wouldn't it mean that you only need X amount of watts only when the load is being lifted?
Besides, I would also need to think about an hydraulic or pneumatic system in order to "soft lock" the mech/exosuit.
After all, if I don't find a way of locking the limbs once they reach the desired position, I would perpetually need energy to keep the limbs in said position.
... Or I could just use a simple ratchet system like used in ratchet wrenchs using really simple electronic mechanisms...
Combustion Turbine Engines:
So, last Project Log I gave crazy and stoopid ideas for a possible turbine engine involving detonation and so on.
With the parameters of the combustion turbine engines, its sizes and its power outputs, it could help us comprehend more or less in a realistic manner the size and output of similar types of engines.
Just like in the Project Log 37, I miss-read the results of the electric motor output and suddenly got all excited on the prospect of powering an entire suit with just 200 watt electric motors.
... But after I noticed my mistake, I got more reallistic results and it helped me "ground" my expectations.
This goes for the turbine engines aswell.
When someone thinks of compactness and high power output, they immediatly think of either a Formula 1 car or a turbine engine (and I don't have money for either).
But let's imagine the output of this "new" engine this way:
For you, dear viewer, this concept may be obvious and extremely easy to just get over with.
... But not for me.
It is the same issue of giving useful advice for other people.
You can clearly see the solution, but I do not, because I can't see through the situation like a third viewer can see.
So, yeah, I need to remind myself: "you can't make a 500 hp engine the size of a backpack with random trash you found on a junk-yard".
One of the reasons why I tried to avoid using combustion turbine engines was precisely the temperature issue on the turbine.
The entire turbine is precisely built in order to create a layer of air flow between the hot part and the walls of the engine; Remember the law of thermodynamics is that everything goes into equilibrium, so even if you have a source of "hot air" and a source of "twice hotter air" mixing together, the result is not a "triply hot air", the "twice hotter air" loses heat to the "hot air".
But in the end, all this heat source goes through the turbine, which has to take all the heat constantly.
The air around the turbine normally reaches 800ºC (1400ºF), steel is literally in liquid form under these conditions.
So, what is normally used is Inconel, a superalloy based on Nickel.
Well, I couldn't find the video, but I remember seeing a video of a guy messing around with an inconel just for the content's sake. But at the end he explains that the Inconel plate he was messing with, that had the size of an Iphone and the thickness of a credit card cost 900$ dollars.
Now you know why jet engines are so fricking expensive.
Anyway, I don't actually need an jet engine, I need a Turboshaft.
In summary, a Turboshaft is just a turbine engine that doesn't produce thrust, only rotational speed. Which is normally useful for helicopters and power generation.
In a super sketchy summary, you don't need as much crazy amount of power as in a jet engine because you don't need to generate thrust (I think).
This means that one could build an entire turboshaft only using cetrifugal compressors... Like in a turbocompressor.
This is a 3D model of a common configuration for a homemade turbine engine, I couldn't find the exact video, but there is a guy that made a lot of these at home (he is an aerospace engineer).
Just like my mood, he starts saying that these kinds of turbine engines are a waste of time and money, lol (I couldn't find the exact video, or I just missed it, bur it is from the guy bellow I think). His one stage turbine engines had a terrible efficiency and a bad compression ratio.
... And it also just produced 8 horse power.
I don't remember him talking about the temperature issue, but since he said that his turboshaft engine had a RPM of 17500 (if I didn't missheard what he said), I would argue that he is limiting the spinning rate and the power output to avoid the engine simply melting.
Well, all of that just to say that there are turbine engines that use water injection, spraying water on the combustion chamber or compressor, so, from the time it reaches the turbine, it already fully converted to steam and avoids the turbine from getting damage.
(planes using water injection during take-off, not the prettiest site to behold >.>)
On top of that, if the water to steam converstion isn't fully achieved, droplets of water will reach the turbine and cause cavitations.
If you don't know what cavitations are, basically, when you hit water really hard and really fast, it creates vacuum bubbles that implode, slowing breaking the surface of metal, like if piranhas were slowly eating it.
I don't really know why the guy in the video didn't use this technique, maybe because you couldn't really condense the steam mixed with fumes back to water and to it again...
But... If he used this technique and added more centrifugal compressors for better fuel efficiency and so on, maybe his turboshaft engine could achieve 10hp or a little more to 15hp.
Schematic of a turbine engine that uses water injector (yes, that spray looking thingie there), although I don't know why inject the water before the regenerator and combustor... There must be a good reason for that, but I don't know which it is.
... But maybe someone could use a second stage free turbine to power a cooling system for the first stage turbine.
This is a turboshaft engine from a company called "GT50" (as you can tell by the watermark).
The number 1 and 2 are connected and are the compressor and turbine of the same engine, however, the part 3 is not connected to the first too parts, so it is basically a free spinning turbine that works with the exaust from the first stage.
So, basically a turbocharger for a turboshaft. Which is ironic, since I would insert a turbocharger in an engine made out of turbochargers... lol
In summary, I could power a cooling system that would create an air flow around the turbine blades of the first stage and avoid meltdown without using water.
I know that I gave the idea of using water, but... If the water will fully turn into steam and be completely mixed with the exhaust gases... Then it wouldn't be really viable to "just" filtrate it... :/
I say this because an electric turbocharger/turbocompressor that runs on 7 kilowatts of energy (around 9 horsepower) achieves around 100 thousand RPM.
And accordingly to the kilowatt/horsepower calculators, if you had this power output and this speed, you would have 0.493 lb/ft (or 0.6684 newton meter, or around 6 kgfcm) of torque.
Turbochargers normally achieve 150k to 200k rpm, and assuming the torque is around the same, you would achieve around 10 to 20 horsepower.
Of course, this is in the best case scenario.
Maybe if you took a bigger turbochargers (that are more expensive) and/or added a heat exchange system before the turbine and use other "tricks" to extract more power, maybe you could achieve 50 horsepower from this turbocharger turbine.
(This is an example of a "new trick" that could be done)
So... All of this trouble just to power the exosuit I talked about in the previous Project Log...
... And in order to power the 1000kg of force Mech, you would need 10 of these working at same time. :|
I tried to search for 400-500 horsepower turbine engines to power the 1000kg mech and I couldn't find many recent turbines, but I did find this turboshaft engine from 1952, which achieved maximum800 horsepower.
I don't think I can copy the insane engineering of a turbine, even though this one was made almost 70 years ago, but I feel like I could copy the size/dimensions of its inside workings.
Also, about the crazy idea for a micro-detonation turbine engine that uses water to damper the detonations and propel the turbine...
Well, like a sane person would expect: I don't really need the water flywheel thingie, I could just shoot the microdetonations in the air.
... And maybe constantly spray water into it without having to worry about cavitations.
Let me explain:
The idea would have a solid rod of metal (such as steel) and drill holes and channels into it, the fuel, air and detonations would travel and exit these channels/holes.
So, from the outside, it would be a solid rod just rotating into its own axis and spitting little jets of fire. Which leads me to assume that it would have the problem of cavitations.
PS: the water spray would also damper the engine noise.
Although I don't know how well such thing will peform.
The closest thing I could thing of was Hero's reaction steam engine.
This is just an example, the original was more or less like this:
Although² I don't really know how I will make this complex myriad of holes and channels throughout the solid piece of metal...
I fear that maybe I will be forced to 3D print this mess and then cast it with metal...
Why?! Why?! Why does 3D print keeps chasing after me?
Of course, all of the things I've been saying lately is "maybe this", "perhaps that", blah blah blah.
Which is not that much factual or scientific, I can't just relly on "maybes", or else I will be just working with hyphoteticals and dreams.
... And that's why I'm trying to calculate the microdetonation impulse turbine engine, because this engine would be so simple to build at home that maybe it could be possible to calculate it more precisely.
The only downside is that I wouldn't be able to use the same tricks as in a conventional turbine engine, such as the regenerative heating and the air flow protection in the combustion chamber... If such thing will have a combustion chamber.
Although I talked big above, the more I think about this micro-detonation crap, the less viable it sounds to me.
I don't even know how to make the combustion/detonation chambers actually detonateinstead of just releasing a bright flame.
I know I said "micro-detonations", but I'm starting to doubt how many detonations per second I will actually need and how well the whole structure will withstand such pressures...
Also, I was thinking again on the idea of just using an conventional hydro-turbine generator and pump the water with said microdetonations, since the water would transfer the detonation more efficiently (I think).
(just imagine the first example from the left to right, but instead of gravity pushing water down a river, it is just micro-detonations pushing water through a tube)
Although I also don't know how well the whole structure will take the micro-detonations and how to make the micro-detonations...
I will try to first find how much force a micro-detonation would have based on the amount of fuel and air present in the mix.
I will try using gasoline as the fuel (although I would like to use Acetylene as the main fuel), simply because it has more information on the engine subject than Acetylene.
It normally has a fuel-to-air ratio of around 14.7:1, which is the ideal mixture, but if you add more air than fuel, then it is what is called a "lean burn". Which is more fuel efficient, but it creates the risk for detonations inside piston engines (which can damage the engine).
... But detonations are exactly what we want. ( ͡° ͜ ʖ ͡° )
Also, accordingly to the video, some engines recicle the exhaust gases in order to cool down the combustion chamber (aka the place where the piston gets pushed by the combustion), this increases efficiency and lowers the amount of toxic gases in the exhaust.
Well, It has been a day or two since I wrote the above matetr, but like I said previously: the more I think about it, the more I feel like I should avoid actually using said micro-detonation engine.
Edit¹:
It was literally yesterday, I'm losing the sense o time, bruh.
How will I inject air and fuel in a rotating micro detonation engine?
The only way I thought was using concentric connection tubes.
Just like in a water pump, there are special bearings and seals that avoid the fluids/gases from scaping, but in summary, I would literally be driving a micro-detonation engine with the fuel and air dangerously close to one another.
If one mistake happens (which will happen), well...
Of course, I could change the design a little, and do the fuel injection from one side and the air supply from the opposite side.
But still... I don't know if I should be doing such dangerous thing...
I mean, I would be inside a fully protected mech (supposedly), and the engine would have a constant injectiong of water to keep it cool, so even if it bursts into flames, the fire would be put would (I think).
But it would be really dangerous to people on the outsideof the mech...
Also, shouldn't I just... You know? Use the combustion chambers to propel the entire engine instead of using the combustion chambers to rotate the turbine?
The combustion chambers are positioned at an angle and at the bottom is the turbine.
Of course, I can imagine that it is easier to rotate a really light turbine rather than entire contraption, which would force the designer to use all sorts of concentric tube connections...
But it would work with a continuous expansion of gases and a continuous layer of air around the combustion chambers avoiding meltdown...
Although the centrifugal forces would force the gases to hit the walls instead of be contained inside the carefully planned structure...
I also remembered about tip-jet engines, basically, these are hollow helicopter/propeller blades rotating at a certain speed.
Due to the centrifugal force, the air and fuel mixture are passively compressed. But even then, these aren't as fuel efficient as the other types of engines and are well known for its insane high noise.
Dunno if it will be useful for me tho... It doesn't have a compressor and I'm not an aerospace engineer to say anything with certainty...
🤔 hum...
This guy made a tip-jet homemade RC helicopter, and bruh...
He ended up being really disappointed with the final result, but this could give an idea on how hard it is to properly balance an entire rotating engine...
In doubt, I think I should just calculate the fuel requirement of this one and for the micro-detonation hydro-turbine idea.
If both end up with similar outputs, I think I will stay with the hydro-turbine.
I remember seeing a video of a guy making a DIY turbine out of wood saw disks supposedly made of "pure tungsten carbide", but I couldn't find the video.
Which would be a really good idea if these saw blades were actually made out of pure tungsten.
Normally it is just the tip that is made out of carbide material... :/
... Plus, normally these are just coated with the abrasive material, not completly made out of it.
Well, I asked ChatGPT to calculate and it gave more or less 400 liters of gasoline per minute in order to achieve 500 horsepower. :/
I know that this AI doesn't know how to make calculations, but so does I, lol.
When I went to calculate more or less the amount of energy contained in each gram of gasoline to convert into power, it was way, way less than what the ChatGPT normally suggests.
And I calculated that even in perfect conditions, it would still consume more or less 60 to 70 liters of gasoline per minute, while normal engines only consume this amount of fuel per hour.
... oof, I forgor to actually write this calculations...
But whatever, I'm not in the mood to calculate this kind of stuff right now, and rellying on an unproven design that somehow would outperform the best engines in the world? hmmm
Very unlikely.
(it took me all this time just to finally convince myself to simply settle down with my stupid ideas, bruh)
And I feel like the idea I had of pumping cool air into the turbine blades would already help things out.
Also, I'm researching some old turbine engines and some times I find this interesting double-centrifugal intake, the air is taken from the middle of the compressor and moved sideways until it reaches the combustion chambers.
I think that this is a neat "trick" to take more air without having a multiple compressors working.
Well, if you know any interesting engineering "tricks" I could use this engine I will make, I would be thankful for your help.
Well, for now, I will try to find the biggest turbocharger I can find and make casted or composite copies of its compressors and maybe its turbines.
I don't really know if composite materials will be able to survive the centrifugal forces, pressures and temperatures inside the compressor (because when you compress a gas too fast, it heats up) or use the previously talked Idea of impulsionating the entire engine with the combustion chambers like a hero's engine.
FulanoDetail
