gravity racing, soapbox racers, gravity cars, gravity bikes, skate luge


You need to log in to create posts and topics.

Scuderia Kouba Mk V build

After a fairly successful rookie year at the 2017 Portland Adult Soapbox Derby (we won with the Mk I and finished 5th with the Mk II), I was unsure of what to do for 2018.  I didn't just want to show back up with the same car and try to take it all again, but didn't have the required motivation to build another one "just because" either.  Then sometime in February or so, the organizers posted that they hadn't had a four wheel car make it to the finals since 2010, never mind win the event.

Challenge accepted.

Last year, we learned a lot about what it takes to build a fast gravity car.  The rules are fairly open (basically less than 5' wide, 12' long, 8' tall; under 300 pounds; less than $500) so you are free to create within those basic parameters.  For the Mk I, we went with about as small of a frontal area as we thought we could manage, 20" wheels in front, single 26" mtb wheel with disk in back.  We applied common sense aero development- make it look streamlined and hope that it is.  We ended up with this as our final race configuration for the Mk I:

The Mk II was similar, but slightly wider and taller based on the dimensions of the driver for whom it was built:


The overall objective for the Mk V was to minimize frontal area and drag while still being large enough to fit me (6' 3", ~190#) and safely convey me to the bottom of the hill.  Based on the results we had from all of our testing the year prior, we stuck with the 20" wheels- this time using them in the back as well and reducing our overall height by a potential 6".  With the max length set at 12', we knew we'd use all of that trying to make a streamlined body.  The Mk I and II were at 4' wide and were plenty stable, so a slightly narrower track was chosen to reduce frontal area but still have turning room for the front tires.  All this pretty much set up the basic layout of the car.  I fabbed up a front and rear axle and set them on the floor to see how things lined up:

Basic components

The next challenge was to figure out how to package it all.  After some basic measurements and figuring out just how wide it needed to be (see the tape on the floor), I shrunk it down again just a little more and started cutting metal to connect the front and rear ends:

One area where we left some speed on the table was at the cockpit.  We assumed the open top design was not ideal, especially in light of the roll structure I built into the rear of the chassis.  I wanted to do an enclosed cockpit on the Mk V.  This would also require planar steering actuation (push-pull) instead of the traditional rotational to keep the steering "wheel" and other components inside the body.  Everything needed to be as compact as possible.  To get a feel of what sort of space was available, a frame for the top of the tail section was tacked in place for evaluation and I climbed in.

I made a bender to arc some 1/2" by 1/2" box tube and started building the tail in earnest.  In conjunction, I bent up a piece to profile the shape to used for the cowling structure, and another for the nose.  I bent matching pieces for the forward and aft profiles of what would become the canopy and connected them with additional tubing.  To get a feel for what it looked like, I started to connect the dots with duct tape.  It at least had the appearance that it might work:

Coming up with the windshield shape was the crux of it all.  I kept staring at conical shapes until I was confident that I could get the shape I needed without having to form a compound curve.  It worked:

With the critical hardpoints fairly committed to at this point, it was on with the rest of the build.  We're allowed a 10 yard push, for which you can use a ~10' 10" pole, so a push point and support structure was built up and more curved pieces were added which would go on to form the body.

Provisions were also included to facilitate a full belly pan for the entire length of the car.

Next up required fabrication of the body.  This began with more tape at the hoops with a bedsheet (budget restrictions) soaked with polyester resin draped over and pulled tight to use as a buck:

Over this, a healthy layer of drywall mud was used to fair in the final shape.  The mud was chosen due to budget restrictions ($10 for a LOT of it!) and it was significantly easier and healthier to use than bondo or other chemical-based fillers.  And a LOT was used to get to the final shape... and then actual fiberglass cloth was used.  In this picture you can see the masking tape through the soaked glass which was used to hold shape while the glass and resin was applied.

Tail section:

The nose was made by spraying insulating foam onto a basic shape and then belt sanding with an 80 grit belt to rough in the shape and then finished with more mud.  Glass was applied over the top of that to form the final shape.  Putting all the raw pieces together, it looked like this:

From here, the bottom of the chassis would be wrapped with a decent thickness of alu to keep pointy things from poking at my squishy bits and the raw glass would eat up an enormous amount of filler (I am a total amateur at this bodywork stuff).  Putting it all together, it started to feel like an actual car.  With the lower chassis yet unpanelled and the nose off, you can see this car is built to be worn like a suit:

With the body basics out of the way, it was time to work on the mechanicals.  One of the design objectives was to have the steering gear out of the airstream.  With this in mind, I set up the front axle with the tie rods directly in the wake of the axle beam.  You can see the general layout in this pic:

The uprights are far from optimal both aerodynamically and kinematically speaking, but they were successfully proven on the Mk I so I went with the known quantity.  Originally I wanted to use brakes in the rear only and to keep the front wheels unadorned with brake mechanisms (reduced aero drag), but in testing, the rear only brakes performed miserably so they were moved to the front.  In keeping with the streamlining theme though, I did internally route the cables (and yes, they are rim brakes- remember, we were on a budget).

So after all the filler and sanding, and with a few cans of paint, things started shaping up:

From here out it was all about making sorting runs and seeking aero tweaks.  We added wheel disks and put airfoils over the front axle.  Eventually we fabbed up some wheel pants and got them hung on the uprights.  Even without them, on our first short course test day, the Mk V was ~3 MPH faster than any of the cars ran there- last year or this year.  Things looked good.

The Mk V in the open

Footage from one of the passes on that day:

We had to address some issues we uncovered during testing- brake mods, steering linkage mods... but nothing which we couldn't overcome with some midnight oil.  Compared to last year, we were cutting it super-close on our timeline.  In fact our first day of testing this year was the same date as our last day of testing last year.  We had it pretty well put together, along with the rest of the fleet, for race day.  The only thing left to do was apply sponsor decals.

So we did:

At the event, the car was extremely fast and handled very well.  The course record is 71.1 seconds and was set on a day where it was 104°(F).  The fastest time last year was set by the Mk I at 75.03 seconds.  The Mk V did exceptionally well, consistently logging runs of 72-73 seconds over the .7 mile course, on a day where it was only in the low 80's.

Spoiler alert: Event results revealed...  It's my intent to (eventually) post on here about the event itself, but I will go on to share that we did win it again with the Mk V, logging a fairly sweet 72.6 second pass for FTOD.  The Mk I fared honorably as well, taking home 3rd overall this year.

EXCELLENT representation of the "backyard" build process. So many different ways for all of us to achieve the same goals.....the beauty of this sport ! Congratulations on the win.

What awesome builds!  Great job.  Everyone loves streamliners.  No doubt they'll haul the mail.  Posts some speeds when possible.  Again, very cool builds

Thank you gentlemen.  I have thoroughly enjoyed the building process on all of the cars and love the optimization challenge our event poses.  We have a .7 mile course with some mild turns at the top, followed by a long straight, followed by a decreasing radius right sweeper entered at your max speed, then it unwinds to straighten out over the finish line.  Too much weight, you burn up energy in the turn; not enough, you get beat down by drag once you're at speed.  It seems like it's all a balancing act.

As for hard numbers, at the event, I saw 51 MPH on the GPS as a max speed.  Reviewing the data file, it looks like that may have been a spike and the real number should be ~46 MPH.  This also matches well with the data from the rest of the runs that day- mid- to high-40's was all we could squeeze out of it.

Importantly, it also handled very well at those speeds going into the big sweeper (called "The Learning Curve" by the organizers).  I was able to make my runs without touching the brakes, entering the Learning Curve at top speed, and tracking well through its entirety.  One of the apps I was using also tracks lateral g's, and I had spikes over 1g and consistent .9g while passing through the apex.  I am pretty impressed by that.

In testing, we have two places where we've gathered data, one slightly steeper than the other.  On the mild run, we were running low forties where the other two cars were high thirties. For the steeper run, I ran low 50's all day long with a top run of 52.3 on that course, which is the fastest I have been in that car.

We were a little bolder last year with the Mk I and Mk II in testing, and located a road with a very moderate gradient but which goes on FOREVER....   We logged a number of 14+ mile passes lasting >20 minutes and topped out over 60 MPH in both cars (62.3 max).  That might not sound too fast in your circles, but the gradient of the road was pretty mellow and we were happy to see those results with our equipment.  Regardless, the overall experience of getting 20 minutes of seat time at one shot was PHENOMENAL.

Before I go there with the Mk V, it'd need a serious braking upgrade.  The rim brakes are "adequate" for derby duty but are not something I'd want to rely on in an open road, top speed situation.


More pics from the day are here:

Our build/testing/event album for all the cars (more pics than you'll ever want to see) is here:



Awesome builds @ckouba


Go Derby!