Weston's (mostly) Scale Model WIKISPEED SGT01

[thingstodo]

Now that I've actually seen one of the SGT01 cars and taken a lot of video and pictures, I feel that I can start on this build.

The plans for the WIKISPEED SGT01 vehicle include drawings of the car's frame, the suspension, steering, brakes, interior, crush structures, and car body.

Link to Open Source Ecology http://opensourceecology.org/wiki/Wi...ikispeed_STG01

Link to pictures from the Winnispeed (Wikispeed build in Winnipeg, Canada) project. The pictures are recent, and are for the installation of the Electric Motor Module into the car. http://lepinsky.com/winnispeedsept2013/index.html


Weston (my grandson) is almost 3 and a half. I'd like to have this working for his fourth birthday (although he likely won't see it until he comes up to the lake, perhaps next May)

I have collected some parts for this build that I think will work pretty well, but I guess we'll see.

Overall:
The full sized version (frame only) is 48 inches wide. This one will be 26 inches wide since I could not fit the parts into 24 inches. Similarly, the full sized frame is 20 inches tall. This one will be 12 for the same reason - I can't get everything to fit. The 10 foot length of the original will be cut in half for the model - 5 feet long.

The model should fit through a doorway. I'd like the turning radius to be tight so that it can be used in a small yard ... we'll see how that goes!

Frame:
- 2 inch aluminum tube, 1/8 inch wall (1/2 size from the 4 inch tube, 1/8 wall in the full size version)
- bolt-together frame (for the model) instead of welded-together frame in the full size version
- motor module will be bolt-together like the frame, using the same methods, but will use 1.25 inch tube just like the full size version. Scaling it down to 5/8 tube did not allow me to use the bolts and washers that the full sized version uses. The larger tube takes up more room, but it should be OK.
- 3/8 grade 8 bolts, with nordloc washers to keep things from getting loose

Parts:
- Recycled 48V Curtis 1204 controller, 300A max (a TINY bit of overkill ...)
- 48V gel-cell pack recycled from an industrial UPS. 7 ah batteries (which will likely put out 3 ah now). I have 3 of these, so they will be set up to charge in 30 minutes (off-board charger to start - there is no room!). Hopefully Weston can make a battery pack last 10 minutes! I should be able to exchange packs and keep the car running for a couple of hours at a time .. I hope. If he likes it, perhaps an investment in a better pack will happen.
- 12V gel cell for auxilliary - LED lights, a horn, perhaps signal lights. Maybe even a small stereo amp to give it a VROOM sound.
- rear suspension, motor and differential from a 3-wheeled old-guy scooter. Rated at 24V but it will work fine at 48V ... for a while
- rear suspension plates cut from aluminum per CAD files ... but the shocks will be modified a bit. I don't have a line on smaller shocks yet.
- front suspension - plates cut from aluminum per CAD files. Perhaps I can modify bicycle rear shocks to fit?
- wheels - recycled from a Carryall electric utility vehicle. 12 inch diameter. Adapter plates to be hand cut for the rear, use the whole hub for the fronts
- steering ... perhaps scale model of WIKISPEED prototype steering rack using non-tempered aluminum? Looking for a small steering wheel ...
- Brakes - front only, part of the Carryall wheels. Cabled to a custom brake pedal
- accelerator - the smallest potbox I can find
- dash - I have some old analog gauges that I may be able to drive with an Arduino. An old-school limited-range volt-meter for the gas gauge, a tachometer to display the rpm from an inductive pickup and a couple of magnets on the axles for speedometer, LEDs for signals and headlights.
- Reverse will use a recycled forward/neutral/reverse manual switch from the Carryall

Unlike my last project - the Electric WIKISPEED Motor Module - I hope to keep this build thread update at least once a week, and to keep it up to date.

We'll see how well that works for me. I have not stopped working on other WIKISPEED projects ... so this may be a bit of a challenge!

[Xist]

"You [shouting at the car] are a child's plaything!"

Oh boy! That is some project! Good luck!

[thingstodo]

I dug out the old-guy scooter parts. There is good news and bad.

Good - I have BOTH a set of rear wheels AND a set of front wheels with the steering linkage between the front wheels. So I don't have to create something.

Bad - the scooter was not nearly as wide as I remembered. The distance between the rear wheels is 17 inches. At 24 inches, the frame is almost OUTSIDE the wheels. I'm not mechanical - so I will need to come up with an alternate method to attach the wheels (both front and back) to the frame. The WIKISPEED design just won't work for this hardware.

My wife saw the parts I had laid out on the lawn and mentioned that the 16 inch width is a bit claustrophobic for Weston even now. And he will likely outgrow it very quickly at that width ...

So 2 votes for '16 inches is just too narrow'.

The 60 inch (half of 120) length seems OK. The 10 inch (half of 20) height looked a bit too low to me originally (plus I had issues fitting the parts in) but I guess I should take another look at it, since I will be likely be raising the frame above the wheels.

I COULD build the frame at 16 inches, with a 10 inch height, and use that as the sub-frame of the car. Then I could build a wider body that bolts to the sub-frame. That solves the issue with 'where to put all this equipment' - the sub-frame will hold it. And that solves the issue with only 16 inches - the upper body can be 24 inches wide, or even 26. That will give Weston (and who ever he's playing with) some elbow room to steer, if I have trouble locating a really small steering wheel. It's manual steering, so the wheel has to be big enough to turn.

In the end, a half-size WIKISPEED car was the inspiration for the build but safety, practical matters, and recycled parts will dictate any changes.

Using scooter parts has another side effect - no brakes. I'm going to have to come up with a method of braking scooter wheels ....

With the narrow track on the steering, no brakes and no suspension, maybe I should not use the scooter parts? I can remove the steering linkage and re-use that part, but use the Carryall parts for everything else ...

[thingstodo]

You have been warned.

Armed with some information (the welded joints need to withstand 500 foot-lb of torque) I can now design a test. It may be destructive, but hopefully not Destructive or not, I need to know.

I have measured torque with a home-built rig in the past, so I have a starting point. Here's what I think I need:
1 - a sturdy wall, that can accommodate connection to the aluminum tube and joint to be tested
2 - a hydraulic jack, to apply the steady pressure to be tested
3 - a scale, to measure the applied pressure
4 - a short section of rebar or something else handy to extend the hydraulic jack up to the joint being tested

The idea is to take a single joint between 2 tubes, fix one tube to the wall at an appropriate height, and orient it to allow testing of the joint at the end of a measured lever. Since the scale only works in one direction, up or positive in the Z direction, the tubes and joint must be moved and re-attached to test a different joint direction. The first test is to bend forward, followed by backward - these use the same mounting holes on the wall. Bend to the passenger side, then to the driver's side requires horizontal mounting of the tube and new holes in the wall. The last test re-uses the first set of mounting holds. Plus an extra piece of aluminum bolted in place to twist the joint at the specified 500 foot-lbs.

Instead of cutting another piece of tube, I connected two of my existing 60 inch lengths of tube with a bolted connection as described in previous updates and videos.

For your reference - Bolt together demo - YouTube Explanation of bolt together frame idea - YouTube

The bolted joint is 3/8 to 1/2 inch from the end of the tube. The contact point for the hydraulic jack is about 36 inches from the center of the bolt, at 36.5 inches from the joint end of the piece. This gives me approx 36 inches or 3 times the weight shown by the bathroom scale.

The analog bathroom scale that I had used for my previous work was apparently recycled through the local Salvation Army. I guess I'll have to invest $20 and get another one. Perhaps I can find a more rugged version at a thrift store? The bed on that scale flexed so I had used a 2 x 8 on top of it to prevent the bed from bending.

The local Value Village comes through with an analog bathroom scale for $6. This looks like the one that I was searching for in my garage .. and could not find. So I am familiar with the ways around the flexing of the bed. Onward!

The 60 inch aluminum tube that has the joint bolted to it has 2 extra holes drilled close to the ends (3/4 and 7/8, since I was not paying attention!). I had planned to use the walls of my garage as the mount points for the test. My wife requested that I 'dig holes' in one of our shed walls instead. I had not really thought of that, but I promised to try. The shed is much smaller in weight and built of smaller dimensional lumber - 2 x 4 instead of 2 x 6 - so I hope the wall does not flex and mess up my test.

I started out using 2 lag bolts 3.5 inch x 5/16 to hold the aluminum tube to my shed wall. The placement of the holes did not allow use of a socket/ratchet combination or my cordless drill. Using a wrench to screw the lag bolts in was very time consuming. And if things go OK than I will be faced with doing this many times ... time for something different. I used a 2 x 4 bolted to the aluminum, and lag bolts through the 2 x 4 to the shed wall. I'm not sure that it will hold, but it's a WHOLE lot easier to put on. If it breaks, I'll come up with something better.

The first position is on the corner of the shed.

Now for some overkill. I am targeting a test of 150 lbs on my scale, at a distance of approximately 4 feet, giving me approx 600 foot-lbs of torque. With this value I am still confident that the joint has withstood 500 foot-lbs. Even if all of my errors add in the same direction. And even if Murphy's Law is followed so that the errors add in the worst possible way. The test value of 150 lbs is also close (2/3) to the calibration weight(me) with my other bathroom scale, so I am cautiously confident that the scale reading is reasonably accurate.

The analog bathroom scale was rough-calibrated with my digital bathroom scale using my weight (223.8 on my digital scale, with my safety gear on, versus 221 read on the analog scale). The analog bathroom scale is within 2 lbs of my digital scale - close enough for me. The hydraulic jack is weighed (25 lbs without the handle) and the scale is reset to 0 with it still on the platform.

Refer to the pictures - 612a tube bolted to a 2 x 4, 2 x 4 lag bolted to the wall, joint sitting low, tube extending 5 feet from the wall. 613a - jack beneath, sitting on bathroom scale. Bathroom scale leveled and zeroed. The top of the jack is a 1.75 inch circle. Distance from the center of the bolt on the bolted connection to the closest contact point of the hydraulic jack is 36 inches.

Set up the video to capture anything exciting.

The way that the shed is set, and the small deck that is attached, makes 3 feet much easier to implement than 4 feet. So I'll need to use 200 lbs at 3 feet instead of the 150 lbs at 4 feet that I listed above.

The pressure is slowly increased - slowly does not really do justice to the pumping that is done to extend the hydraulic jack. Somewhere around 7 pumps per 5 lbs pressure. At 50 lbs, I notice that the joint is no longer square but do not dig into it. After raising the pressure to 66.6 lbs (200 lbs torque), the joint deforms enough to have the weight on the scale drop off. There is some close-up video of the joint. More close-ups to show where it deformed - the middle wall of the tube mounted on the wall, where the bolts go through. The shed did not move. The lag bolts held, at least for the limited torque exerted so far. With this failure, I see no reason to proceed with the other directions of the test.

After unloading the joint, the square shows that most of the movement was permanent. 5 mm was measured as permanent. there may have been an additional 1mm that was elastic.

Video - (long and boring, good resolution so it takes a while to download!) Update Saskatoon Sep 23, 2013 - YouTube

What to do from here?
Since the deformation was in the middle of the tube wall, I guess I should avoid that area for bolting in the future. With the 2 inch tube that I am using for Weston`s truck (his parents cautioned me that a car was not `cool`. All of the toys for little boys are trucks, or tractors, or quads. So the model SGT01 frame will be modified slightly to look more like a truck) .. where was I? Oh - I don`t think that the 2 inch tube (1.75 inch inside dimension) has the room for 4 of the 3/8 bolts side by side. 4 * 3/8 or 0.375 = 1.5 inches ... which leaves a total of 0.25 inches clearance between the bolts. That divides down to 0.083 inches (1/12 of an inch) per hole drilled for alignment issues, not drilled plumb and square, marking accuracy ... significantly tighter than my present construction methods will tolerate!

Perhaps I can use two bolts (instead of the one I used in this test) to connect each of the 1.75 inch tubes to the main tube, and a single bolt (instead of the 2 I used in this test) to connect the second tube at the joint? That should minimize any distortion on the side wall of the main tube, as happened in this test.

What about the 1.25 inch tube being used for the motor module? 2 bolts from 1 direction and a third from the other direction adds up (3 * 3/8) to more than the 1 inch size. Perhaps a single bolt could be used with a washer to reinforce the wall of the aluminum tube?

Looking ahead to the 4 inch tube, moving the bolts as far as possible to the edge of the tube wall, and using 2 bolts per 3.75 inch tube (for a total of 6) should again minimize the problem with distortion on the unsupported edge of the tube. Will it be enough? Who knows? We`ll see!

For the near future, after I get the changes implemented on the 2 inch tube and they successfully pass the 500 foot-lb test, the 4 inch tube will be next! it may take a couple of tries to get the 2 inch joint to pass the test. That's one of the reasons why I'm testing on smaller and easier to handle joints.

Whew! If you're still reading, I'm impressed!

[thingstodo]

From the September 23 update - what I think I learned:

1. Using one bolt to connect an open square tube to a joint does not work well
2. It takes time to gather tools and parts to test bolted joints
3. My shed may not look very sturdy, but it will likely do OK for testing up to 500 foot-lbs of torque on bolted joints .. perhaps more ..

For our next test joint ...

The 1.75 inch long 3/4 inch tube that makes up part of the joint will have a hole drilled in each end close to the end. The 9/16 head of the 3/8 bolt should jam against the edge, so the hole should be just under 5/8 from the edge. This should assist with assembly.

The 1.75 inch long 1 inch tube will have similar holes drilled.

Pictures of caliper, punch marks from end of piece, finished pieces

These 2 tubes each 1.75 inch long will be bolted to a new 60 inch piece of aluminum tube. The original 'second' 60 inch tube will be bolted to create a joint that looks like the original test joint, but should withstand much more pressure .. I hope.

The new joint will be bolted to 2 new 2 x 4 pieces. These pieces will be lag bolted to the shed at a height of 15 inches, like the September 23 update.

It's raining and it's dark (I got home late ..) so the testing will have to wait until the weekend.

I know this was not much progress .. but smaller chunks should be better than the last marathon update, right?

[thingstodo]

The bolt-together joint was built.

As described in the previous post, the new pieces were aligned on new pieces of 2 inch tube. 4 holes were drilled through into the 2 inch tube and bolts were installed, threads facing out. There was room to install lock washers this time.

The second 2 inch tube was pounded into place (still a very tight fit). A single hole was drilled through and a 3/8 by 2.5 inch bolt put through to secure the tube in place. I may not be getting better at this, but I'm getting faster!

The tube was mounted on the shed (not quite vertical, as it turns out ... but close). The same equipment was used to test:
- 2 x 6 base
- analog bathroom scale
- 2 x 4 on the scale to prevent denting the deck
- 20 ton hydraulic jack
- square to check deflection of the test joint
- caliper to measure deflection

I took video. The first is a summary

http://youtu.be/gsqUvlhQ-R0

The next 4 are exhaustive video so that someone could reproduce the tests if they were so inclined. I don't recommend watching it if you are easily bored. 4 videos so that it fits into the Youtube maximums.

http://youtu.be/TSndXAVtQr0
http://youtu.be/Aj0R2x9KMus
http://youtu.be/XAzm_fGinrg
http://youtu.be/q2EblcLIDsY

The tests show that the tubing I am using - 2 inch square 6061-T6 aluminum - will not withstand the required 500 foot-lbs of torque. The joint flexes out of square almost immediately (30 lbs force) but the flex or deformation accelerates above 300 lbs force.

The failure, or deformation, occurs in the 1/8 inch wall of the tube that is bolted to the wall. That is the main tube of the car frame, the part that everything else is attached to.

The tube wall thickness could be increased, but increasing the wall thickness of this tube would substantially increase the weight of the frame.

That portion of the tube wall could be reinforced with angle or a welded element that is annealed in an oven to restore it's temper. That requires a skilled aluminum welder and an oven with a control system to restore the aluminum tempering. And the additional reinforcing would block access to the tube, where access is required for bolts, wiring, etc.

For Weston's vehicle, I expect that this frame will be sufficient for a few reasons:
- Weston weighs 30 lbs, and I am designing for his uncle to ride with him in a modified seat, behind Weston and sitting atop the 'motor module'. Total weight 300 lbs. Note that maximum speed will be GREATLY reduced wit this much weight on the vehicle!
- Each joint can withstand 300 lbs. The weight in each axis will be distributed across at least 4 joints on the front and 4 joints on the back
- The failure of the joints, when they deform permanently, is based on the imposed force. And it is not a catastrophic failure. The 90 degree joint was tested to 300 lbs of force. I was expecting the Aluminum to 'tear' but I did not reach that point. At the speed that this vehicle will be limited to, the tires will go far out of alignment but the structure will not fall to the ground at a dangerous speed.
- There will be a full roll cage for Weston (not for an adult riding behind him, though). I have not sourced a 5 point harness or a seat to accommodate that as yet.
- I still have the option of adding another 1/8 inch of aluminum reinforcing on the inside of the 2 inch tube, for a bit of extra design factor
- That's enough design factor for me until I can do more testing!

[thingstodo]

Not much of an update this week

There has been discussion on how to reinforce the bolt-together joint so that it will withstand 500 foot-lbs without distorting .. or at least so that the joint will take the torque without failing.

The suggestions are:
1 - use a large washer to prevent the pull-through that was experienced
2 - use epoxy to bond the short pieces of tube to the longer main tube. This is the weakest point, and could also be bolted to prevent catastrophic failure in the case where the epoxy fails. I know nothing about epoxy beyond using JB weld for many years.
3 - use a welded joint, and test it the same way to see if 500 foot-lbs is a reasonable amount of torque to be requiring, or if the weld has much more strength than that.

The washer idea has had some progress. I cut a 1.5 inch by 1.5 inch square of 1/8 inch steel from some scrap that I have. I have not gotten the holes drilled yet. The aluminum tube used will be new (not recycled from other testing)

I was pointed toward some epoxy, 3M 08115 or 08115, Plexus combined with PC-120 primer, Epovex by Zyvex (can't find much info on that one). So far, no one local sells these. I need to ask the suppliers what they would advise, I guess.

One of the forums suggested JB Weld - but I think that was for an aluminum boat - it may be interesting to make one up with JB Weld for comparison, but I think it makes more sense to do the serious testing first.

[thingstodo]

There has been progress on preparing for the next test of a bolted joint:

1 - I cut out 3 plates measuring 1.75 inches x 1.75 inches from 1/8 inch aluminum stock as well as 1 plate 1.75 inches x 1.75 inches from 1/8 steel stock. These will work as 'large washers' to (hopefully) prevent the bolt heads from pulling through the wall of the aluminum tube. This is the failure mode that I have seen in testing so far.

2 - I picked up some epoxy.

The aluminum needs to be prepared similar for both:
- cleaned with alcohol
- sand off aluminum oxide
- clean off the grit

Loctite 330, from one of our local vendors that handles the Loctite product line. This is intended for metal, quite simple to apply - activator on one side, adhesive to the other, assemble within 5 minutes, clamp. 24 - 72 hours cure time. 2000+ psi shear or tension.

Permatex Cold Weld metal epoxy. A 2-part epoxy that claims 4500 psi strength ... not sure whether that is tension or shear. Mix equal parts thoroughly. Coat both sides and clamp. Sets up in 8 minutes.

3 - I dropped off 5 pieces of tube and they have been butt-welded at 90 degrees to form 3 joints. I will be picking these up on Monday and testing them. Perhaps to failure, perhaps to the extent that my equipment will go.
I messed up a bit - I don't have any 2 inch tubes left that are long enough to allow testing with 1 (bolted) or 2 (epoxy) methods.

So .. the plan is to test the welded joints from part 3 on Monday. If the joints fail, I'll cut the welds off. If not, I'll rotate the tube 90 degrees.

Epoxy the joints on one end with part 2 and drill the other end with the plates in part 1.

Testing on the bolted joint can be done on Monday if time permits, but not on the epoxy joints as they take time to set. It will likely be Saturday before I can test them.

But these are just plans - we'll see what I get done

[thingstodo]

This is not so much an update, as a project split.

Building a scale model of the WIKISPEED SGT01 is where I started the project. Things have changed along the way:
- cars aren't COOL - it needs to be a TRUCK
- the frame is too wide for the wheel assemblies that I am recycling so changes/additions need to be made to the frame
- I could not fit the electric parts into a true scale model, so the measurements on the basic frame were changed
- the axles being used don't conform to the WIKISPEED motor module requirements, so changes are required there
- since the axles don't work as they should, the suspension modules need to be modified

Sounds like this build won't be very close to a WIKISPEED model, right?

And I'm not a mechanical kind of guy. I'm doing mechanical because it needs to be done, I enjoy the electrical parts of the build.

ssssSSSSSSOOOOOOooooo ...

I've decided to do an ACTUAL scale model build of the SGT01 WIKISPEED car first. This may end up being a bit sparse - no body or interior perhaps - but it will have the correct measurements for the FRAME, the MOTOR MODULE, the SUSPENSION, and the various CRUSH STRUCTURES.

There will be lots of Plywood used in place of actual suspension components, as in 'dummy' shock absorbers, maybe the A arms, etc. But I will be able to SEE and TOUCH a scale model that measures out correctly and has all the correct proportions BEFORE I make what appears to be MANY design changes to convert a CAR to a TRUCK.

Since I am lazy, I will be asking one of the local fabricators here, Marine Master, to to supply and cut the aluminum pieces for me. I plan to put these pieces together with bolts and epoxy (based on the upcoming tests that I will be doing on Monday).

Now to justify my laziness .. their shop charge is reasonable, and it will likely take them about an hour to do what I need. AND doing this will save me 2 or 3 weekends of metal cutting and filing (not fun tasks). AND it will look better when they are done than I could possibly have accomplished no matter how much time I put into it.

I am planning to get the joint assembly process done this week after I get my testing on the welds done, so I can make all of my mistakes there. It usually takes longer than I expect, so that will likely take 2 weeks. After that, when the aluminum pieces have been cut, I will take pictures and video of the assembly process. It should be instructional even if it is scaled down.

The updates here are still Weston's build. The information from the joint testing will still be here, as well as my progress on Weston's Truck.

I will start a new build thread for the WIKISPEED scale model SGT01 when I get enough progress to actually report on it.

[thingstodo]

The project continues.

My tangent - testing how to bolt together the aluminum frame instead of welding it - continues.


Update from Oct 30

Demo Oct 30, Welded Joint testing summary - YouTube

Update from Nov 7 - Summary posted to WIKISPEED channel tonight (forgot to post earlier!)

Demo Oct 30, Welded Joint testing summary - YouTube

Update Nov 11

Demo Nov 11, Joint Testing V3 part 5 - YouTube

Uploaded to the wikispeed youtube channel

Demo Nov 13, Joint Testing V3 part 6 Wrapup - YouTube


I expect that any solution that I implement will be a combination of bolts, large washers, and an epoxy of some sort.

I expect that I will be busy with the testing for another month or so