Basement Racer

I began the real design work by building the Basement Racer. The Basement Racer is a full-scale mock-up of the formula car. I have a set of Miata wheels standing in position where they will be installed on the finished car. I also have a driver's seat with adjustable back and seat. The seat sits on a pair of 2x4 beams set up like frame rails on the floor. An additional piece of plywood supports the engine.

This page is organized chronologically. The most recent material is at the bottom of the page. My goal has been to show how the concept developed and why.

Vehicle Shape

With the very basics in position, I began working on the computer with engine dimensions and drivetrain layouts to position all the hardware pieces of the car. The complexity of all the parts though made it difficult to get something on screen that looked reasonable and my results always looked too complicated.

I already had some PVC pipe on-hand to mock-up real frame tubing, so I went and bought some 2x4s and framing screws to help begin giving the car shape. I built a forward frame and a rear frame on the car to use for forming my PVC main tubes. I used additional screws to set in as pegs into my frames so the height of the PVC bows could be adjusted at the forward, middle, and back of the car. A couple hours of building and bending allowed my progress to leapfrog forward.

Of note, I have discovered why so many formula cars look a bit chubby. The engine forces a minimum practical width for removal. My engine will lift clear of the car with a hoist. With the 20" max engine width and the curved tubing tapering in width aft of the seat back, the seat back is 24 inches wide. On the upside extra width is extra space for fuel behind the seat. Formula Fords and others with inline fours along a longitudinal axis are actually narrower too than this motorcycle engine when the bike engine is mounted transversely. I could narrow the car four inches fairly easy if I turned the engine, but that would prevent the use of chain drive so I'd need a transaxle. Given my project budget, that's not an option.

The nose may also become a challenge although I plan to work my way through by prototyping in cardboard and more PVC. The short nose will place the double wishbone suspension right at the front end of the car. I had also hoped to place a radiator up there either horizontal or slightly upward canted with air flowing up and through for cooling. It might still be acieved, but the packaging will require significant effort and sidepod radiator mounts may just make more sense.

At the tail, I have PVC tubing hanging out aft of where I had sketched the back end of the car originally, but the tapered tail looks very good like the back end of the boat-tailed Alpha Spyder or like the Ariel Atom. The extra length also opens up some more involved drivetrain options at the back like retrofitting a transaxle later.

Seating Position

A notable design feature visible in the older pictures above, is that the footbox for the driver may be raised. Raising the footbox leaves space for a nose diffuser like those seen on modern prototypes like the Audi R8 (www.mulsannescorner.com). Raising the footbox though significantly impacts the aesthetics of the car and visibility over the nose.

Seen here, there is a significant difference with the raised foot position versus the low position. Given the narrow cockpit and body of the car possible with a single seat, I'm likely to drop the raised nose option to make the car more aggressive looking without the need for full bodywork. The narrow nose leaves adequate space on either side for respectable downforce generating elements...

On the other hand, the steering wheel makes visibility over my toes a secondary problem. A 12" wheel looks like the largest size that I could practically fit in the car, and a 250mm (9.84") or 200mm (7.87") wheel would make things that much easier. An easy-off wheel would be my first choice, but I need to find out if that's legal for street use or if I'd need more permanent attachment for legality on the road. I may need to change the steering wheel after arrival at the track.

The Post-Atom - Agressive Structural Concept

If I step up to my current status, there is a significant shift. While my frame concept up to this point clearly resembles an Ariel Atom, it's more a path of parallel evolution than direct imitation. Going on a side-trip, you'll notice that the four-tube frame that I defined to flow about the cabin actually bears sigificant resemblance to a curvy Reynard Formula Ford sized for one person or any other number of similar frames based on just four axial tubes connected to one another. Even before these frame types were common in cars, they could be found in simple aircraft. Look half-way down this page for pictures of the wooden "four-tube" fuselage structure:

http://www.clifdawson.ca/Pietenpol2.html

One of the weak links in this design though is that the truss cannot easily resist torsion loads. Torsional or twisting stiffness of a truss frame requires triangulation between at least three tubes and the third tube must not be in the same geometric plane as the other two tubes. This isn't too hard in the tail of an airplane, but things get more difficult when you get to the cockpit. In either an airplane or a car, the section of the truss where the driver goes, triangulation across three tubes is not possible without placing the driver outside the structure or placing a tube through the driver. This is less than ideal since the inside of the vehicle's structure is usually the safe place to be in a crash and removing te driver is important between races. The choice then with four tubes is safety or stiffness.

Several auto companies found ways to deal with this challenge though in the 1950s and 60s. Mercedes made the 300SL with a deep side-sill for a triangulated truss with some out-of-plane depth to improve torsional stiffness. This stiffer frame served both the streetgoing sportscar and the touring racecars very well. Maserati took a similar but more extreme approach with their Birdcage racecar. The Lotus Elan took a different approach, placing the driver and passenger outside of the main bending and torsion structure.

http://www.seriouswheels.com/mno/Mercedes-Benz-300SL-Gullwing-Coupe-Drawing-1920x1440.htm
http://www.atspeedimages.com/pebble2000/laguna/maserati_birdcage.jpg
http://www.atspeedimages.com/pebble2000/laguna/maserati_birdcage_behind_dashboard.jpg
http://www.lotuselan.net/publish/ttr_frame_mods.shtml

Since the key is three tubes then, my car has picked up a third tube on each side of the driver. This allows a triangulated truss on each side of the driver to provide bending and torsion stiffness.

When all three tubes on a side of the frame are triangulated (Triangular Truss), the car will gain significant torsional stiffness and the added depth of the car's side walls will give a significant crash protection bonus. The cost is increased weight that the car will carry for the benefit of extra stiffness, and a significant increase in complexity and number of tubes in the frame. This structure will probably take twice as long to build as a four-tube frame. That's the reason that a lot of cars stick to four tubes. Most skip to aluminum or composite semi-monocoque rather than deal with the large number of tubes and welds. That means that mine will be the only one on the block though.

Rollover Protection

With the general frame developed below the shoulder line, I took a look at another significant structural concern. A roll bar is a must if the car is ever to compete, and given the agressive nature of the vehicle, it's a good idea in general. The roll bar shown above rises to a peak height of 48 inches, the same total height as a Mazda Miata. With that rollbar height and the driver in full upright seating position for comfort, the rollbar is still a bit skimpy for competition rules about clearance over the helmet. 48 inches is the height while the car sits with 5 inches of ground clearance. I anticipate that the car could be lowered to 2-3 inches of clearance for racing which reduces the height of the rollbar by 2 or 3 inches, but that still leaves the bar four inches higher than typical for a comparable DSR class racer like the Cheetah (http://sports.racer.net/chassis/cheetah/gallery1.htm).

The reality though is that there are two paths of development that I'm trying to mesh. There's the street legal machine that might get some cross-country miles, and there's the formula car for A-modified or full-bodied DSR racer if I go club racing. In most cases, racecars make poor streetcars and the same goes for cars that try to go the other way. The tall roll cage though happens to be a poor compromise for both cars. It's too large and gaudy for most street use, and it's taller than desireable for racing (love how I call the rollbar gaudy when I'm building a formula car). Both missions say the same thing in this case: make the rollbar lower. I've lowered the bar by six inches to 42 total height for now and I'm taking my time to decide what I think. 42 inches on the road and 40 inches height at the track may be a good setting. The broader solution to make the rollbar work is to use different seat trays for racing and for street use.

The street chair will keep the driver more upright and comfortable for longer journeys. This will leave less rollbar height margin with the driver's helmet essentially as high as the rollbar, but this shouldn't be a problem when driving at 5 to 7 tenths. For racing, the seat pan will be lower to keep the rollbar several inches overhead. This will cramp the legs tighter and reduce over-the-nose visibility, but it will keep the car's frontal area more modest. Crunching the driver into a space smaller than what's comfortable for long periods is a key to keeping the wheelbase short on many racecars. If I held a racecar comfort standard, I could shrink my wheelbase 5 inches. Since I also plan to drive this on some weekdays, I'll keep the 95 inch wheelbase. The Lotus 49 has a similar wheelbase but took five or more inches from the driver to give that space to the engine and transaxle it appears.

Cockpit Canopy?

Everyone likes a sleek bubble canopy. I might try making one for this vehicle at some point. The roofline peak forward of the car's mid-length is a really cool signature of mid-engine cars too.

Latest Progress

Development work on a project like this often becomes a stop and go affair. The need to save money for major tools and parts though can be a blessing, providing time to really think aout design decisions - or to agonize about even the littlest details. The picture below though shows how a bit of time to think about the design has changed the outboard frame tubes. They now arc down to a middle-height finish rather than pushing up above the other frame tubes as seen in the pictures above. This revision places the car's structure closer to the rear end differential and suspension hardpoints. It also makes low bodywork more convenient where the raised tubing would have forced increased vehicle frontal area.

The next pair of pictures below show the bulkheads at the front of the driver compartment for suspension pick-ups. With the taper of the nose up-front, the hole pattern for the front suspension will not be the same for both bulkheads. The best method for handling the suspension pick-ups as they angle in toward a convergence point in front of the car may invlove some specialty machined fittings. A secondary benefit of this approach would be that the fittings could be replaced to change the suspension geometry of the car.

The real shock to the front bulkhead process was test fitting of the forward bulkhead. The rear bulkhead took a couple hours of cutting, fitting, measuring, and re-trimming to go into place. For the front bulkhead, the tubing was all positioned around the rear bulkhead and measured. After cutting each slot for tubing, everything went together on the first try. Working on a 3D collection of parts with a ruler and a few pieces of wood, getting a perfect fit on the first try has been a rarity.

Besides physical framing work at the front of the car, the rear has been developing. These two pictures show a plywood piece that sits flush on the floor pan of the car and connects to the engine through two of the engine's original mounting bolts. These two bolts will be used to cantilever the differential case off the back of the powerplant unit at a height of 12 inches above ground. The plywood plate is oversize at the moment and will be trimmed back. The major challenge will be how to route rear suspension bulkheads around the differential and chain drive assembly. The forward bulkhead for the rear suspension will be roughly seven inches forward of the differential and driveshaft assemblies. This will place it a couple inches behind the engine. The goal with my front and rear suspension bulkheads will be to use side plates to close in the bulkheads like six-sided cubes to keep all the suspension hard-points square to each other. With luck, this will help to ensure accurate anti-dive, anti-squat and precise wishbone positioning.

As a final shot , I have the view over the nose of the car. The major problem with this photo is that the car is in the basement facing a white wall. It makes it difficult to tell how high o low the road will be out over the nose. A Wyle E. Coyote style tunnel painted on the wall might help, but I'll just wait until things are in a state for a proper design review up on the driveway.

Suspension First Assembly

At this point, it becomes more critical to get the suspension geometry in place so work can commence on triangulation of frame tubes. I have been converging on a solution involving front and rear sub-frames for the suspension. The frames are machined as plate pieces, assembled, and then used as self-jigged tooling to be welded into the car's frame. The front and rear subframes will coordintate the positions of the main axial ubes of the frame with the goal of making a car with minimal warping in shape.

The front sub-frame was finished last night and the a-arms were cut to size today.

Likewise, work is underway at the rear of the mock-up to frame in the rear sub-frame. This one may be more challenging because its large size makes it more valuable to pursure weight savings. The large sides of the box will also be susceptible to warping, oil canning, and other deflections. There are good odds that this rear box will be built as a set of picture frame sides with large side openings and external triangulation between the frame tubes will be used to provide torsional stiffness for the rear of the car.

Notice that the rear engine mounts are about 1.5 inches from the forward bulkhead of the rear sub-frame. The rear engine mounts would be a logical place to link the engine back the the car's structure, or to mount a frame for the differential. At the moment, I lean toward making an opening in the bulkhead behind the engine and and mounting a differential frame to the attach points. It may be practical to extend the differential frame aft much like the frame at the back of a West DSR, but in this case the rear of the frame can be attached back to the car's frame. I want to make provision for either hard or soft mountings back to the car. If it becomes an issue though, the engine and differential will be hard mounted.

Frame Features Revised and Systems Integration Work 4/8/07

With a good deal of the frame geometry defined, it becomes important to look at the next items that may cause the design to stumble. I need real-estate for the brake master cylinders and sterring system at the nose of the car. The brake and clutch pedals were spaced and positioned originally to bolt right up to the rear wall of the front suspension sub-frame. Given my intended driver size range though, I need axial adjustment for the pedals. This would preclude bolting the pedals to the bulkhead with the master cylinders on the far side. To get the full range of adjustment without pushing the suspension forward, I need to turn the front suspension sub-frame into a tubing structure. Likewise, I find at the rear that it could be very difficult to box in the entire sub-frame box. I have not given up on an enclosed box concept for the front and rear suspension, but given the required tooling and the workload on all of the Wichita area machine shops this year, it just may not be practical.

To keep moving forward, the key is to begin filling in the blanks that I do know. The steering rack is in rough position at the nose of the car and I have installed a mock steering column. Getting in and out with the steering wheel nstalled would be a task. A removable wheel will be important.

Take a look at the cardboard behind the steering wheel in the pictures above. This will be an additional frame tube or a built-up dash which will serve both as a roll-over structure and a mounting point for insruments. I have also found that the cross-member at the front of the cockpit opening is a bit close. This cross-member will be moved forward and the built up dash will be built on the top of the cross member. If I make a built-up dash, the built unit will replace the cross member tube to alleviate local space congestion. Triangulating members in the sidepod structures will move with the dash cross-member.

An additional topic of discussion is pedal setup and shifting. I'm told that most of the bike transmissions are stout enough that many of the FSAE teams do not clutch to shift except for first gear. Space is at a premium in the footwell. Placing both a gas and brake pedal on the right of the steering column would require narrow shoes. Brake on the left and gas on the right without a clutch pedal is much easier to package.

Speaking again with former FSAE team members or people who followed the cars, a steering wheel clutch handle or some other reasonable position seems attractive. As an alternative to having a clutch actuator at all, there have been some interesting electric and pneumatic shifter methods. One interesting method used electric buttons for up and down shifts on the steering wheel. These energized a clutch actuator, and when the clutch was depressed fully, a limit switch would energize a circuit to make the shift. Setup I'm told took some effort, but the result is the opportunity to remove an un-necessary control from the cockpit and significantly drop shift times. Looking at an acceleration chart for an automobile with manual shifting versus a bike with no-lift shifting quickly illustrates that manual shifting takes more time. At this point, I'm going to plan on not having a clutch pedal. If I do install a manual clutch control, it will not be foot operated.