Thursday, March 29, 2007
Tuesday, March 27, 2007
Here's a short movie of the beveled gears turning. It is more encouraging than it looks at first. I think it will work just fine with some slight adjustments, a little realigning here and there. I think
I will get the T joint working and then add a fourth shaft.
Monday, March 26, 2007
So to test some ideas for the output shaft I pulled out my old three-gear test and inserted another copy of the large gear. Here the pinion gear, large gear and "shaft" are turning on rotary jigs. I was hoping I could use the sulfur atoms on the end of the gears as teeth, looks like I can. By keeping the output shaft and output gear separate pieces I might be able to fulfill a grand vision of using a linear motor to build a type of clutch that engages the output shaft to the output gear. Now I will build the actual output shaft by extending a copy of the large gear slightly and placing the teeth at the end of the gear. Or I was thinking of trying to make some beveled gears. That would look pretty cool.
Sunday, March 25, 2007
Last night I fell asleep to the pitter patter of raindrops and the whirl of cooling fans. When I woke up this morning the rain was still going, but the whisper of the fans had stop. Lately that has instilled something of a Pavlovian response in me because I know it signals the end of a simulation. Click the pic to see the final carbon nanotube speed reduction box. Lowering the temperature didn't really do that much for the wobble. However I ended up placing a few anchor jigs around the perimeter of the casing and that seems to help a lot. So what's next? Well, the original plan for the box included a shaft on the output gear ending in a T-junction. I'm thinking I might pursue that.
Friday, March 23, 2007
Click the pic to see the cnt speed reduction box do a little dance. This is a pretty short clip, but I have a much longer one running too. In all honestly this one is not rolling at room temperature. What? Liquid nitrogen is really easy to get these days. In addition to looking a little drunk, there appears to be some slippage on the large output gear. I am not sure if that is just a quirk of the animation or what, have to wait for the longer simulation to finish to find out. Also, for the long simulation I lowered the temperature even more to see what that does to the wobble. What? Liquid helium isn't exactly the rarest stuff in the universe either.
Wednesday, March 21, 2007
Not much to say, click the pic. I only have half the casing in this simulation. I think that was a mistake. I should of used the entire casing for stability and just have hidden the half facing the viewer. To me it looks like the casing is what ends up breaking apart the large gear. I am going to run a simulation on just the gears and see what happens.
Sunday, March 18, 2007
Here is an animation of the pinion gear turning the middle gear on the cnt bushing, click the pic to see it in motion. For this simulation I used the anchor jigs to secure the bushing. In the finished box the bushing will be bonded to the casing. The middle gear has quite a bit of travel on the bushing; I might need to adjust that later on. Next I will rebuild the large output gear with the new spacer designs.
If you click on the picture of this bearing and shaft it will open a new window with some animation of it turning. Right now I have a simulation of the pinion gear, middle gear and cnt bushing running on the other computer. I have a feeling this version will be more stable. When it's finished I am thinking about posting an animated gif of it; however I am not sure if I made the simulation long enough to be interesting.
Friday, March 16, 2007
I hope that is my last Charlton Heston reference, but it is keeping with the agony and the ecstasy. I was spending a lot of time on the gear re-design for the CNT SGR without ever really being happy with it, specifically the spacers. Originally I had built SiC rings and, more or less, haphazardly attached them to the ends of the tubes to keep the gears aligned in the casing. I have thrown them out in favor of a much simpler design which is supposed to be the ideal of engineering anyway. This time I attached one Si atom to each carbon atom on outside rings. The Si atoms were then bonded to each other to form couplets. On the outside of each couplet I deposited a sulfur atom, and on the inside hydrogen. This is a much simpler spacer and looks like it will be more stable. Above is the new pinion gear without the teeth added yet. I am going to build the other gears in a similar way and see how they work out.
Tuesday, March 13, 2007
Well the first simulation of the CNT SRG did not go ideally. This may be because of a small bug in the simulator that I have a patch to fix. Also I was looking at the gears with the casing hidden, and it looks like they are no longer all the round, so I am trying to fix that before running another simulation. Lastly I am toning down the rotor torque that drives the pinion gear and lowering the final speed by half. I will run some smaller simulations on just the gears until I get that working. These simulations do not take nearly as long.
Monday, March 12, 2007
After a little over 88 hours the simulation is 89% complete. It should be done tomorrow morning. The total atom count for the CNT SRG is 11,914. In the mean time I have been working on another CNT speed reducer, trying out some different designs and techniques. I am not sure how well it will work, but what I have done so far is posted above. This is more of an experiment in connecting CNTs of different diameters using SiC junction rings. These rings are a lot like the spacers I used to align the CNT gears in the speed reducing box below. However this time extra carbon atoms are deposited on the inner circumference to decrease the inner diameter. The idea is the inner carbon atoms will bond to a small diameter nanotube and the carbon atoms on the outer circumference will bond to a larger nanotube, joining the tubes together. If this doesn't work I will try using a heterojunction CNT. This is a single tube that changes in diameter along the length of the tube. However within the software I don't think I have the ability to choose the two diameters. This is why I am trying to develop techniques for joining any two nanotubes of different diameters together. Ultimately I would like to build some kind of mechanical calculator, perhaps like Babbage's Difference Engine. One has already been built out of Legos. If I am going to do this alone though I will definitely need to up grade my computer power.
Friday, March 9, 2007
Looks like I was wrong in my intial estimate of 30 hours. I am bumping that up to 100. Also I was so excited to start the simulation that I started it with the gear box in a weird perspective, so after this simulation I am going to run it again with a better view. In the mean time here is a picture of the planetary gear with half of the casing removed. This allows you to see the sun and planet gears inside.
Thursday, March 8, 2007
I finished the casing and attached the bearing for the middle gear today. I started the simulation calculations; my guess is they will take around a day for my computer to crunch through. After that I am going to make an animated gif file that shows the box in motion, assuming it doesn't fly apart or anything. This should be at a lot higher resolution. I'll post it here when it is finished. I just did a little arithmetic; at this rate it will take around 30 hours for the simulation to finish...
Wednesday, March 7, 2007
My speed reduction box is coming along. I got some major help with the gear case. It is silicon carbide. I started trying to build one out of diamond even though SiC was recommended; my first case looked pretty crappy. This case is not done yet but gives a good idea of what the final look will be. I had to rebuild the drive and output gears to fit it. I used the same design for the outer sleeve portion of the small bearing as spacers to keep the drive and output gears aligned in the case. The middle gear bearing will be attached directly to the gear case. I'll be working on finishing the case next.
Monday, March 5, 2007
Here are the three gears for my speed reduction box. The gear on the left is the drive gear that powers the other two. The middle gear uses a second nanotube as a bushing. The largest is the output gear that, hopefully, will drive a couple other CNT shafts. The drive and output gears will pass through the gear box casing, and the bushing for the middle gear will attach to the casing wall. I might need to make the bushing a little smaller in diameter and extend the output gear a little. I was able to set up a second computer today, so I had parts cooking in a couple pots. More to come...
Sunday, March 4, 2007
This is a planetary gear box. This one has eight planet gears. It has taken the better part of the day to make, but, again, a lot of that was waiting for my computer to make the computations. I worked from a tutorial, but ran into several problems. There is a lot of hidden structure inside the outside casing; if I get around to it I will post some pics of the sun and planet gears inside. The main problem I had was with building the sun gear. Something went wrong and I had to manually bond every other oxygen atom on the inside of the gear. Also I accidentally deleted an entire row of carbon atoms but didn't notice until a while later. I had to piece that back together.
This is about as complex as I can get with my Toshiba laptop, but I think it is six years old. I am going to look around and see what I can put together in terms of a faster computer. Next project might be a semi-original speed reducer using carbon nanotubes. I will try to sketch out a complete design on paper first before I try to model it in NanoEngineer-1. I am thinking I will post it piecewise here, so you can follow along with the agony and the ecstasy.
Saturday, March 3, 2007
This was my first attempt at building something without a tutorial. I chose Nano-Velcro . I've tried to model it in the past, with much to be desired. I came much closer this time, but, as the second image shows, mine wants to untwist at the end of the simulation. This is because Nano-Velcro uses pentagon and heptagon deformaties in the carbon lattice to make the hook part. I started off with a 20 A CNT (m =0, n=7) and manually bonded a second prefab CNT with the same chiralty, but this one I bent 180 degrees to form the hook. This only creates a bent CNT and not a rigid torus like in the original design. I changed up the default colors so it was easier to see the separate tubes.
Nano-Velcro has been proposed as a very strong, self-repairing, way to mechanically bond stuff together. As an undergrad I wrote a hypothetical business plan for a company using it to replace wire bonding in the production of integrated circuits. After the class presentation I found out one of the other students had accepted a position with Applied Materials, one of the largest producers of IC-manufacturing equipment. Maybe I should of insisted on a non-disclosure agreement?
Here I used another tutorial to recreate the Nasa gear. Two carbon nanotube were inserted into the workspace with a chirality of n = 16, m = 0 (actually only one is inserted, the other is copied from it). The chirality describes the asymmetry of the carbon latice. For a more in depth explanation of carbon nanotubes check out Physical Properties of Carbon Nanotubes by R. Saito et. al. These are formed into gear mechanisms by "transmuting" two carbon atoms at regular intervals around the tube from sp2 to sp3. Four additional C-sp2 atoms are added to each transmuted couplet to create a hexagonal tooth. The remaining unbonded carbons were then hydrogenated. The tubes were then placed into an interlocking position. Again building the gears really didn't take too long, but running the simulation gave my outmoded laptop a work out. I think I am going to have to break out my other rig, much more crunching power there.
Above is a picture of a molecular bearing I constructed using a tutorial for NanoEngineer-1. The entire process took about an hour, but I imagine much of that was due to learning the software and the speed of my trusty old Toshiba Laptop. In the future I am sure I can get better renderings and even animated simulations. The atom break down is as follows:
White - Hydrogen
Yellow - Sulfur
Black - Silicon
Gray - Carbon
Red - Oxygen
The idea is the inner sleeve, encased in sulfur, is allowed to rotate inside of the silicon/carbon outer ring, just like...well, a bearing. I will go into greater detail on the functioning of these machines in the future. More posts to follow...