Having a busy week at work so there's not much evening activity going on cart wise. We did spend a couple hours last night adding the composites to the center chassis former.

It was going to be a bit of a pain in the ass to use the carbon tow in this short space so we grabbed a few yard of unidirectional carbon fabric and wrapped with that. We again finished over that with one layer of 5.8oz glass to protect the brittle carbon.

Next up ... I'm going to weld up and install the front steering mechanism and wheel support. On it's own this will add some stiffness to the front portion of the chassis and after it's installation we will determine if the front former is just stiff enough, or needs composite layers added, or can be removed completely.

We build a set of crude headstocks so we could wrap the prop pylons by spinning them. It was a ton easier this way rather than looping the spool around and around like we were forced to do on the chassis itself.

Here's Rick starting the windings.

100 layers of carbon tow on each pylon and a layer of 5.8oz glass later:


Before the wraps we did a torsion test and recorded the results. Once this cures I'll publish the pics showing the before and after tests. Hopefully we added a good bit of stiffness.

We received the 'correct' steering bearings today.

Cheap Chinese bearings are easy on the wallet but sure can be a pain in the a** sometimes. These are supposed to be for a 3/4" shaft, but as you can see the .738 ID is going to require me to do a bit of material removal.

Oh well, 5 minutes work with the grinder and we'll have a .737 steering shaft. :-)

Tested the sample of Qpower line that we received. Stretch and strength as advertised and as expected. So far the stretch characteristics of Kevlar, Spectra/Dyneema and Vectran have all be similar - well under 1% under our proposed loads.

This product has a thin woven sheath and I feared that this would add a bit of 'squishy' to the end result. I was pleased to find that under load, I could do the usual gentle caliper measurement and then squeeze very hard and only get a .001 or .002 difference. Pretty happy with that as the sheath will up our durability - though it does add another layer on the reel due to the diameter increase.

I'm pretty sure that this line will be the line that we use to test the spool transmission for suitability.

The students had a presentation due today relating to the project. I dropped in to catch it in person and on video.

After their 15 minute presentation, they were grilled by other students and the Professor regarding various aspects of the project.

There was a bit of stage fright going on and I can certainly relate.

The bearings arrived for the front steering mechanism. I was waiting for these before welding up the front steering fork and mounting brackets.

I ordered the right bearings and this was reflected on the packing slip --- but they sent the wrong ones. These things happen.

Should have the replacements in by Friday.

Holloween Workothon

A few thousand wraps of carbon tow, a bit of of glass fabric on the bias and a quart or so of epoxy resin later ...

Holloween Workothon

Foam on and ready for carbon/glass wrap.

Holloween Workothon

I worked on cutting and tacking up the parts for the front forks and steering mechanism. The bearings should arrive today so I'm going to wait for them before welding up everything for good. Check twice, weld once as they say.

Holloween Workothon

Dat and Aung worked on cutting, sanding and fitting up the foam to the chassis pieces.

Holloween Workothon

With our layup plan now in place, some of the gang showed up Saturday morning to help with the chores.

Chris helps Rick disassemble the frame in prep for lamination...

... and ends up at the whiteboard talking transmission theory (or some other brainteaser ...with Rick you never know).

It was testing time for the chassis beam samples that we laminated a few days ago.

First the bare control sample: it weighs 4.5lbs and ~9 degrees of twist. That's a 30lb weight out 24" on the arm.


Next the fiberglass sample: 5.5lbs and ~4 degrees of twist

Finally the fiberglass/carbon combo: 5.7lbs and ~2 degrees twist.

So, it's settled ... will will use the fiberglass and carbon combo. 5.8oz cloth and 70 layers of 12k carbon tow wrapped diagonally.

.

For those of you wondering why we're using the labor intensive tow rather than using carbon fabric, it is mostly because we *have* the tow already -- we paid $60 for 16,000ft of it as potential tranmission line, but it failed for that purpose. Rather than spend near $500 for carbon fabric we will do a little hand wrap work and save both money and weight.

Going in for the easy layup

Aung, Dat, Sheetal and I were up till around midnight last night laying up the torsional test samples. It was quite an adventure as we had to MacGyver a hot wire rig together. Since we have quite a bit of foam cutting ahead, today we'll work on putting together a better wire setup. Aung and I cut this foam by pulling on a wire between us while Dat bravely held the wire to the terminals of an RC starter battery.

The one on the right has just one layer of 3.7oz glass and the one on the left has about 20 diagonal wraps of carbon tow under the same layer of glass. The control is in the center.

Saturday morning they will be cured and we'll test against the control. I'm betting there won't be a dramatic difference between the two and that neither will be stiff enough without another layer. But then, what the heck do I know?

We're drillin' for oil

We had our weekly team design meeting today and just so everyone could get a feel for the scale of what we were doing I clamped all the primary chassis pieces together for viewing. When they imagined the tip of the prop another 10ft above the prop pylon, they were better able to see why the cart has the width and length it does. More than one commented that it looked like an oil drilling derrick.

The rear axle cage that you see here is only for dyno testing -- the one used out on the runway will be twice as wide. The prop stand you see there is 12ft high and is sized for a 20ft prop. Those tall pylons will be faired out to a long airfoil shape for low drag. Even at only windspeed, those pylons get a lot of airflow from the wash of the prop so they need to be sleek.

Bob Parks was at our meeting and over the next few days will be helping us determine if we will be building a 20ft prop or one as small as 16ft turning a bit faster. The larger prop is more efficient on it's own, but gearing losses on a slower turning prop can erase those gains. He's going to help us find the most productive balance.

Before we laminate the foam/glass/carbon on our chassis members, we want to build a test fixture and run samples to optimize the layout for best torsional rigidity vs weight.

We will save one of these pictured samples as a control and layup the other three in varying combinations of glass and carbon. We will then test the three for the best stiffness/weight ratio and against the control to see if any of the three give us the improvement we need. If none of the sample are rigid enough, we will go another round of composites over the existing layers and test again.

We hope to have these samples/tests done by the end of this weekend so we can then start layup on the chassis proper.

Film at 11

We've got the crazy idea that as we get closer to the end of the project and towards the actual demonstration of the cart we might get someone from say the Science Channel interested into turning it into a segment or entire episode. With this in mind we have taken to recording a lot of footage of the design and build process. With this footage in the can, we believe it will be a lot easier to get someone interested at that later date.

Fortunately in our business we can get our hands on some nice video equipment that's laying around. Unfortunately we don't have anyone to actually run that equipment while we work so we just move these around on the tripods and hope for the best.

A generous contribution and big thanks

After load testing our carbon tubes, and settling on a larger propeller than originally intended, we concluded that the carbon tubes we got from Joby were plenty strong, but slightly more flexible than we’d ideally like. Consequently, we put the word out to the local windsurfing community to see if anyone had any carbon windsurfing masts they were looking to sell cheap.

Most carbon windsurfing masts come in two pieces, and typically the bottom piece is the portion that fails. Fortunately, we’re looking to build our prop blades around the upper portion. We were lucky enough to get a call from local windsurfer and friend Jerry Bertrand. Jerry actually donated 4 complete carbon masts for the project.

We did a bit of matching and load testing on these, and determined that two will be ideal for our prop blades, and the other two will certainly come in handy for honing our skills in working with composites as we practice making prop blades. I suspect they’ll also make their way into the project or test fixtures before it’s all over.

Thanks very much Jerry.

The right stuff

Checking pilot position after a long day/night and a full head of welding helmet hair.

All those wood pieces you see will now get the foam and fiber/carbon treatment for torsional rigidity. We'll glue 1" foam against the webs, run a hotwire along the wood to reduce the foam to the appropriate thickness and then wrap first with glass cloth, and then with carbon and again glass for additional stiffness as needed.

It's been weighing on us

Stacked the primary frame components up on the scale to see how we're doing so far -- 140lbs

Every little bit counts

Aung removed another 10lbs of fat just from the webs of the primary chassis rails and formers. Add this to the tapering and cap trimming and that's about more than 40lbs just from those areas.

The frame on a diet

Team members Aung and Dat came over on Sunday to help out with the Mule. Aung spent most of the afternoon and evening continuing with the removal of material where it is not needed. From the start we have removed almost 70lbs of wood from the core materials that we bought.

Moving to the rear

Here I've started making the rear axle/transmission enclosure frame.


This frame is 6ft wide and will stay this width for dyno testing etc. When we take the cart outside, it is designed to accept a plywood box beam built around it that will double it's width for stability.

Demolition derby

One of the primary purposes this cart we call "the Mule" will serve is as a platform for testing different components. There has been a lot of discussion regarding which wheels/tires will both provide low rolling and aero resistance and also be strong enough.

We would like to use cheap and available bicycle wheels, but we'd be way overloading them and also placing on them sideloads for which they are not designed. Buuuut ... if they will take those loads, they sure would be light and easy.

In the spirit of the test mule, we've decided to just throw some on, load them up with sandbags and abuse the hell out of them and see what happens -- science at work.

We wanted a small diameter BMX fork/wheel for the front so we went on Craigslist and found one for $40. When we showed up, the front wheel was so loose and wobbly that we just headed back to the car until the guy hollered "Will you take it away for $20?".

Rick spent some time taking both the front hub and the headset apart and installed a new bearing kit in each ($13). He then hacked the front off the bike and we will weld this into a frame at the nose of the cart.

We will turn the fork aound, mount the headset vertical and likely increase the castor from existing. We hope to steer with our feet from simple pegs sticking out from the axles.

Rick wants a linkage and lever to steer with his right hand. I've told him I'll build whatever he wants if he's in the shop from 7am to 10am to show me exactly what that is. I'm quite safe as he doesn't even get out of bed by then on weekends. ;-) Pegs it will be.

Hidden hinges

The upper 75% of the prop pylon tower will hinge forward so we can cartop travel the assembly. We are going to fair the pylons themselves so these hinges needed to be internal to the pylon spar caps.

The pylons rise upwards at an angle (leaning towards each other for triangle stability), but the hinge pivot axis needs to be horizontal. Embedding a straight piece of steel into the cap put each end of the hinge too close to the edge of the cap for good strength. I welded up an offset hinge that would not weaken the cap too much.

The hinge material will stay vertical and the top cap will pivot on a large pin. I will radius the bottom of that top cap to allow it to pivot cleanly.

Rick at his best

Rick and I both layed between the primary forward chassis members and we marked our knee/waist/elbow/shoulder/head locations along the rails. We need to put formers in the frame and want to locate them for as broad a range of pilot heights as possible.

When too much isn't a good thing

Just a shot to show the material (foreground) that these tapered and trimmed members were tapered and trimmed from. As you'll see in the next few entries, we spent a lot of time putting the frame materials on a diet.

The (rough) shape of things to come

Had a heck of a good weekend working on the cart chassis. During the last week or so we had worked on tapering the front chassis members and working on methods to increase torsional rigidity.

Here's a shot of the central steel box frame with the front members and the lower prop pylon stanchions clamped in place. The upper pylon stanchions will sit upon the lowers and will hinge forward 90d for travel.

Kevlar line test

Yesterday Bob Parks gave us a sample of some Kevlar thread that he thought might work in our spool transmission. It's got no sheath, is slightly twisted and is rated at 400lbs breaking strength -- sounds almost perfect.



I pull tested this Kevlar using approximately the same methodology described in the Specra test. With the lower rated strength of this line I only tested to 300lbs rather than the 400lbs for the Spectra. I did not test to ultimate break as I wanted to return Bob's sample to him intact if possible. I will do that test with my own sample if we feel this material holds promise.

The stretch characteristics of this line were almost identical to those of the Specta in our tests -- ~1% in our load range. This is pretty much as expected. Spectra does have more 'creep' over time under load, but our duration will be rather short so I don't see that as much of a factor.

With a bit more research into availability and price, we should be able to settle on a line to use in our test mule.

Downwind faster than the Daedalus

We had long lunch with the legendary Bob Parks yesterday. Bob is Chief Designer for Aurora Flight Sciences (http://www.aurora.aero/) and is also well known for his involvement along with Mark Drela, Juan Cruz and several others as primary participants and designers in the Daedalus project. This adventure was documented in the book by Gary Dorsey "The fullness of wings" -- a exceptionally great read for design geeks and others alike.

Here's a shot of that record breaking human powered craft in the air:

Bob, in spite of all the various demands on his time has generously jumped in with both feet to help us ensure the success of this relatively small (by comparison) project. His knowledge is invaluable in all areas of the venture, but he is particularly focusing on idealizing a prop design and also consulting on the transmission components. The transmission of the Daedalus was his primary responsibility and thus he brings a load of experience in lightweight and efficient transfer of power to the project. We are very fortunate to have him involved.

From all of us to you Bob -- Thanks.

Spectra stretch results

Just tested that Spectra Qline we got from Rick. Didn't include any pics because it's the same test rig as seen a few posts ago and this line is the color of the floor so there wasn't much to see anyway.

This line is rated to ~600lb and once I got to 400lbs I didn't see the need to break it and ruin any more line than necessary. We plan on being in the 2-250lbs range on the spool drive.

Methodology:

I wanted any slack in the fixtures to be taken up, so I pulled line to 400lbs and then released the pressure down to 100lbs. I marked the line and referenced it against the floor and then pulled it up to 250lbs and checked the reference markers and repeated at 400lbs.

Stretch:
100lbs to 250lbs = 0.7%
100lbs to 400lbs = 1.5%

Below 100lbs I could not be sure where the biggest 'stretch' results came from (chain catenary, etc) so I leave that out -- additionally, I plan on prewinding the prop spool on snug to make sure there is no slack in the loaded spool.

All in all, the stretch testing came out better than I expected for Spectra and within the stretch that I was hoping for with the carbon. This line is far thicker than the carbon so 'spool stack' could introduce some energy robbing compression into the system. We'll have to do further tests with the actual spool to know.

JB

Alternate lines

From Rick:

I spoke to the folks that make the Q-Powerline I use for kitesurfing today. They spent about half hour giving me all the pros and cons of various materials and structures of line.

They're willing to sell me some custom Q-Powerline that should be ideal for our application I would think. It's supposed to have a 700+ lb yield strength. It has an unbraided spectra core, with a very fine braid spectra sheath. The stretch is quite minimal, and it will be very supple compared to the normal Q-Powerline as they've agreed to skip the step of impregnating the braid. For our purposes, we should never come close to the life cycle of this stuff.

Sparring before the big fight

We did a bit of flex load testing on our potential spar material this evening. We tested both the carbon tube that our sponsor Joby Energy contributed to the project and the top section of a carbon windsurfing mast from Rick's garage.

The carbon mast is tapered and significantly larger diameter than the 15/16" carbon tube and to the surprise of no one was about twice as stiff. We'll have to wait on our prop design to know which spar will fit inside the airfoil.

The load you see is quite close to the load that will be on the prop in action. Per Bob Parks, we also want to see the spars handle 2x working load without failure, but this will require building some sort of fixture to hold the supported end -- if we did that with the rough rig you see in the picture we would likely just crush the carbon at the pivot point, ruin the tube and learn nothing. One more test fixture to build.

Got a line on that?

After the failure of the carbon tow as a suitable material, we are going to test a few alternatives.

Rick came up with the Spectra/Dyneema "Qline"(pic below) that is supposed to test out at ~600lbs. We'll see. He's also going to find us some braided Spectra kite line of about the same strength, but without the sheath that is included in the Qline. This sheath adds weight and stiffness that we don't want.

Bob Parks of Daedalus fame is also consulting on the project (how cool is that?) and he just emailed that he has some braided Kevlar that we can test.

In addition to the 6" testing fixtures seen in the previous pics, I also built 2", 3", 4" and 5" fixtures. Once we settle on a particular line, we will test on the progressively small fixtures to see what gearing options will be available.

Stubbed Tow

As previously posted, we got a reel of carbon tow (yarn) to test on our spool drive. I was skeptical (but hopeful) regarding the strength specs from the manufacturer. We built a couple test fixtures and hooked them to the tension guage. The tow is wrapped around the 6" tubing about 20 times and the loose end is clamped under the mini-clamp.

The results were FAR below the specs -- the tow repeatedly broke, not at the anchor radius (where I would have expected), but simply in the 15' strand being tested. Average strength over repeated tests ... 75-80lbs. A far cry from the 560lbs spec.

We'll test Spectra/Dyneema next.