Monday, December 14, 2009

UGV work

Another kind of cool development is that we were asked to see if we could make a unmanned ground vehicle,UGV, with the same brains as we use in the UAVs. That combined with our ripe fruit turret idea gave me a bunch of new ideas about how to separate the hardware from the software stacks. The turret will use some interesting ideas on color detection and correction to detect whether fruit is ripe enough to be picked.

Now there is a UGV in the mix. What better a system to tow the turret around, than a 1/5th scale 4WD buggy. It is also leading us onto a bunch of other applications with the same device. Our new ideas, we hope are very investment worthy. If President O actually does something for small business, then we may be able to attract some investment. I think that there are several interesting campus or property management applications that will make the software easy enough to develop and/or repurpose.

Hopefully, I will get some more pictures of our electric conversion of the buggy and some assembly stuff. So that everyone can get an idea of how to do the electric conversion and the other steps that we took setting things up. We have some math that we have already worked out to hopefully help reduce some of the initial setup mistakes. Many scientific papers all say that small alignment errors in setup can ruin the mathematical assumptions that are made in software. That is why we keep trying to set up our software and then back out the math. Not always a working approach, but it is better than thinking everything is perfect.

Neah, engineers are never perfect... we are within 20%, 80% of the time.

As If You Thought I Have Not Been Working

So I have been trying to get some interesting pictures of the layup of the Big Baby fuselage. However, the pictures that I do have are not as demonstrative as I would like. I guess I will have to get down there and take some myself. All in all, the mold concept worked and the internal, integral structure also seems to be pretty successful.

Our idea, is to mold a certain amount of internal structure into our composite molds. The interest is to reduce production time of aircraft to the limit of curing and finishing time. More advanced resins could be used in the future. Our experiments are with normal 60 and 90min set air-cure epoxies. We also used a male mold to reduce our total investment. Surface finish is less than important to see for these experiments. They are about the viability of integral structure composite lay ups. I have said that last bit a few times, but I have received several comments about this.

The next step is to relief the wings so that they make a pocket, lift system that will effectively create a similar cross-section to the metal equivalent airframe.

Thursday, December 3, 2009

Adventures in Composites: A Bright Idea

So after many months of gnawing on how to reduce the construction time of uavs and model aircraft, we finally decided to watch a million youtube movies. Sating ourselves on anything that had anything to do with composite lay ups we set off onto our own project. The first version of this idea is a proof of concept experiment rather than a real product. Aircraft Spruce sent us some great almost silken 2.5oz/yd^2 fiberglass.

This is an experiment in how to create multi-cell,internal structure in composite structures. If an airframe just needs a few reinforcements, that would go a long way to reducing the production costs of a unit model. Having built several models now with varying levels of success. It is clear, the longer it takes to produce a model, the more likely it will have issues. It could be as simple as being heavy, due to liberal adhesive use. Or just general issues such as fiber breakage or delaminations due to mis drilling or toughness issues in large unsupported panels.

After watching many others' techniques and lots of research into how successful airframes are created. It is about reducing the handling of each piece. The best paint jobs are ruined by finger prints, finishes scratched, just because they had the most chances to get ruined before delivery. Our concept centered around simple low curvature lay ups, that could be lain individual. Then stiffeners installed before the final layers are put down. Each mold would then be set to cure. Final assembly, would be put in place in the planes perpendicular to the molds to join the sides together and to increase the technical stiffness of the airframe.

Our design required a center web and some easy radii transitioning each of the perpendicular webs. We sectioned our fuselage to have three dorsal access bays. The fuselage was then bisected so that we had two symmetric halves and three continuous doors. Empennages and noses are difficult to build and to shape. There is always an issue with the symmetry or the strain in the materials during the curing process. Careful attention was paid to each of these areas to reduce the known issues that had ruined some of our previous projects.

Our split molds were then mounted to three inch styrofoam wall insulation. Not the ball stuff, this is the small,closed-cell material with wood glue. Everything was painted several times with latex enamel. The enamel did not stick during the first few coats. It probably has to do with the static charge that built up during the work. There was an eventual critical coating area that finally whetted the surface. Once the surface finally whetted, the coats proceeded quickly.

We found several movies about hand-lain composites that kind of tipped us off on how not to complicate our chemistry. They all had a few things in common, latex paint and paste wax as the sealant and release agents. After fourth coat of latex paint, Johnson's paste wax was applied liberally and hand-buffed to a shine. Each mold and the foam several inches (3-4") from the model was covered and buffed.

Then it was all let rest for a day.

Adventures in Composites: Two Bogans and the Art of the Lay... Up

After a year of sitting, our polyester resin had spoiled. So we contacted some people and they said you could do it with just about any medium cure time epoxy. Out into the proper disposal container for petro-chemical waste. Yeah, I went there. Off to the local hardware shop for some epoxy. We bought several batches of 60 min epoxy. Some of the information that our friends told us suggested that we could use isopropyl alcohol to help keep the viscosity of the epoxy down as it is laid up.

Information like this is when it gets exciting. We found out a few things. You can paint the alcohol onto a layer to help flow the epoxy and to reduce the bubbles and flatten wrinkles. I would suggest that pour some alcohol into a cup and paint it on where you need it. Each layup will be different so it is hard to say where you would have issues.

There were some bubble generators in the seam where the semi-mold met the backing. This is probably pretty obvious, but it was due to training air under the resin laden cloth. Our mold also had several regions with nearly perpendicular sections which were hard to wet. In areas like that, I would suggest liberally pre-painting. Be careful not to pool the resin in the bottom since that is not where you want the part to be strong. Another observation was to make sure that when each layer is put down, both people should work in the same direction. Do not work the resin from the nose and the tail. It makes wrinkles and other defects.

It is very important to make sure that each layer has as much of the resin out, or at least even distributed as possible. This means that all of the bubbles, wrinkles and puddles need to be worked toward the edge of the layup. A roller and brushes are the weapons of choice for this work. The handle edges are great for wringing out excess resin from tight radii.