Grassroots Motorsports

An Idea Takes Hold Jan 9, 2006

There I was staring at the lonely CSP Miata sitting on jackstands in my garage acting as a storage bin for various automotive objects, contemplating my next move in getting it back in action. The motor had been removed and disassembled after the 2003 season in preparation for a full internal SP build (.047 overbore, balance & blueprint) yet I never seemed to get around to taking it to the machine shop. I had been consumed for two years by an STS Civic build-up and the excitement and cameraderie of the burgeoning STS class. But that car was nearing the end of its development cycle, and like any engineer, I was starting to run out of ideas and getting a little bored with it. So, what to do?

In addition to being a veteran Solo competitor, I also serve on the Solo Events Board and am one of two board liaisons to the Street Touring Advisory Committee. I was there when the idea of the STS2 class was hatched and played a key role in constructing its ruleset. At the time, I believed strongly that the target cars (CRX, early Miata, early MR2, early RX7) would all be contenders, with the Miata and CRX likely to have the most aspirational support from the autocross community. In short, those are cars people like to autocross and they are cheap to setup and race. Yet somehow, the CRX has shot way ahead of the Miata in terms of results and now performance expectations. This has turned STS2 into "yet another Honda class" and I believe is hurting participation. A strong rivalry would be better. How to change that?

A couple of other factors played a role in the decision that followed. I've personally always loved the way a Miata handles. I'm just "at home" in these cars and its one of the reasons I built mine in the first place. I first ran it in ES (nee CS) for two seasons with a number of national event wins, and a fourth place finish at the Solo II Nationals (fast time in class on one course!). But it was no match for the MR2 so I built it into a screaming fast CSP car. Oh my!! Two successful years of campaigning that car had brought it to the point of needing a bit more power to get to the front of the pack consistently. But my co-driver wife starting suffering from neck and back problems resulting from a car accident and the hard-charging Miata just beat the crap out of her while driving. The G forces that the CSP Miata would pull both laterally and accelerating/braking are simply astounding. So we toned it down a bit by building the STS Civic and running on street tires.

With all that running through my head, the idea hit me...why not combine all of that and built up an STS2 Miata? Mazda pays contingency for Tours and Pro Solos so potential funds would be there to help finance the effort. It would be a car that we could drive on the street again between events, and would be a wonderful touring vehicle for drives through the mountains and along the beach roads as we travel the country racing. Furthermore, by doing so "out in the open", the entire Miata community could be engaged in the effort so as to share information and jumpstart the efforts of everyone (note: the Honda community is much better at this than the Miata folks are). With this case-study knowledge readily available and some excitement generated by the public development effort, it should help build STS2 participation in the hopes of gaining national recognition for the class. In short, we could do for STS2 what the early STS Civic pioneers (i.e. Jason Tipple and Chris Shenefield) did to finally kick up interest in that class. And my wife Ann could get what she wanted...damned car out of the garage!!

Planning:

Any good development effort requires a plan, with updates done along the way as new information comes along. The goal is now pretty clear: Build a nationally competitive Miata for the STS2 class as an "out in the open" case-study, engaging the Miata community to increase participation in STS2 for national class recognition.

So who's "we"? This kind of big job sounds like team effort. In order to provide the largest possible voice for the project, a quick call to our good friends at Grassroots Motorsports Magazine yielded their support via this forum and some potential print exposure. Ann & I have known Tim, Per and JG from way back in the early days of "AutoX" magazine and find that the publication has grown and adapted nicely to become a unique force in the motorsports community, especially among autocrossers.

Secondly, we'll need to farm out certain areas of the car's development that cannot be handled in-house. One of those is the motor. More on this later, but suffice it to say that there are few engine builders that I trust to build a fast and legal motor for solo, and Ed Gilfus at Applied Racing Technology is one of them. Ed has been part of the Austin-based racing crowd for many years, spending time chasing cones, driving in circles and a couple of stints in road racing. Like many, he has gotten caught up in the Spec Miata bandwagon and together with partner Karl Zimmerman formed a business around that opportunity. More info on their backgrounds, products and services can be found at www.appliedracingtechnology.com Oh yeah, did I mention that they have an in-house dyno?

Another key element once the car is actually built is alignment and tire service. I've yet to find anyone as meticulous and accommodating as Bill Kim at Soulspeed. They've been a sponsor of ours for the last year and have state of the art equipment for alignment and tire mounting. In short, I wouldn't trust anyone else with my lightweight rims or to produce a repeatable and precise alignment. Bill can also do suspension installations and recommend racing or street performance setups. Check ‘em out.

And while they are not a specific partner in our project, Mazdaspeed Motorsports Development (www.mazdamotorsports.com) deserves mention since their support frees up funds and reduces costs through their contingency program and racer parts discounts. If you are racing a Mazda, "Mazda Comp" is your friend!

Starting Point:

Since the ST rules inherit most of the suspension allowances from Street Prepared, we have no major development work there other than some potential retuning of spring/bar rates, and alignment to accommodate the switch from R-compounds to street tires. We also have two years' worth of experience with street tire testing for the Civic that will apply directly to the Miata. In fact, the wheels/tires from that car will bolt right up!

Because ST has no update/backdate like SP does, the key "undo" will be the motor swap and engine management. In CSP trim, the car sported a '99 1.8 motor with a set of TWM independent throttle bodies and TEC3 engine management. That will all have to go. We'll need to source a complete1.6 motor (the car is a '92 model) and get it freshened. We'll also need to swap out the super-lightweight kart seats (SP-legal) and get some ST-legal (and infinitely more comfortable) racing seats. Seatbelts and steering wheel have to go back to original, as well. And finally, the clutch-diff rear end will have to go in favor of an OE viscous LSD. The latter are tricky to find in good condition and are no longer available from Mazda.

As far as unique performance parts, we'll need a 1.6 header, some sort of intake, and an engine management solution. Some of these will require significant development efforts to maximize performance since there are not good off-the-shelf solutions as there are in the Honda world.

Timeline:

The stretch goal will be to have the car together and sorted enough to take it to California for the first Pro Solo in March. Much of that will depend on the engine rebuild timeline since this is the time of year when all the Spec Miata folks are having new motors built and ART is very busy. The rest of the car is very straightforward once it has been dug out of the garage. From there, we'll do dyno work to optimize the power output and setup testing at some practice sessions. Car should be "done" by June. Of course, I thought my first Civic build would take a month and it took twelve.

And so it begins...

--Andy Hollis

PS: Better formatting and some pics to follow as I figure out this blog setup...

Future topics

• Intake testing
• Header testing
• ECU development
• Tire testing
• Suspension tuning 

Weight Analysis Jan 10, 2006

Another key part of the decision-making process for this endeavor is determining competitiveness of the car. In this installment, I'll go through some of the analysis that shows why a Miata should be at least as good as a CRX and establish some performance-related goals for our project.

My experience has shown that one good measure of potential autocross handling performance is the weight-to-tire ratio. Given equal weights, the car with more rubber on the ground is faster, and given equal tires, the lighter car is faster. Since the CRX and Miata have similar gearing in 2nd, and since that gearing provides a perfect range of speeds for typical autocross courses, both optimally use the same size tires.

The 205/50-15 size is the de facto standard and will be our initial choice. Since STS rules allow up to a 225 width tire, I tested the wider/taller/heavier 215/45-16 size last summer and found that they were only marginally faster on a pure handling course, yet were consistently several tenths slower on a course with some acceleration zones. This result was likely due to the gearing and weight disadvantage of the bigger tire. That said, in the near future we'll be testing the 205/40-16 size that Falken makes in their new RT-615 line. It's shorter and lighter, so it could potentially provide a couple of tenths better performance. The only issue on that tire is that its tread width is not quite as large as the 205/50-15. The only way to know which factor dominates is to test.

With tires being an equal factor between the CRX and Miata, let's now look at weight. Our '92 Miata weighed 2015 in racing trim when it ran in E Stock. Let's work forwards from there to get an estimate of STS2 weight.

Wheel/tire ---------------------------------->(22.5 - 30) * 4 = 30 lbs added! Seats (we'll use stock rails on one seat) 60 - 35 = 25 saved Exhaust (Borla -> straight pipe) -------->16 - 4 = 12
Crank pulley --------------------------------> 5 - 1 = 4 Springs (lightweight 2.25")---------------->15 - 9 = 6 Perches (aluminum) ----------------------->10.5 - 4 = 6.5 Battery (sealed motorcycle) ------------->21.5 - 5 = 16.5 Misc (bushings, intake, sway bar, etc.) ->9

That's a total weight loss of 49 lbs which puts us at 1966 for a target. Not bad, and if the 205/40-16 tires work out, they are 5 lbs lighter per tire which comes out to a total weight savings of 16 lbs when you account for the slightly heavier 16" rim. That combo will put us right at 1950.

So, what does an STS2 CRX weigh? Well, a fully prepped STS Civic can be right around 2000 lbs and its Honda sales brochure curb weight is 2161. The curb weight on an 88 CRX Si (the lightest year) is 2017, for a difference of 144. That would put the best fully-prepped CRX at 1856. So the Miata is giving up a full 100 lbs to the CRX! Wow!

But wait! There's more! The weight distribution of an STS Civic is about 60/40. Its likely to be even worse for a CRX since the leverage of the long tail section is gone and the front end is the same. Let's say 65/35. In ES and CSP trim, the Miata was around 53/47 so that's unlikely to change much for STS2. In a pure steady state condition, with equal sized tires all around, the front tires will saturate first and will be the gating factor on lateral acceleration (and thus, speed). With 65% of its weight on the front wheels, the CRX splits 1200 lbs between its front tires. The Miata splits 53% of 1950 or 1033 lbs. Hmmm...advantage Miata!!

In addition to the pure steady state performance, the Miata's drive wheels are at the rear. As the car accelerates off a turn, the undersaturated rear tires have some additional tractive capacity available for forward bite. Thus, power can be added earlier in the corner exit phase. On the other hand, the poor FWD CRX must wait to accelerate until the turn ends, the steering wheel is unwound and some traction is available to accelerate with. Oh, and did I mention that the Miata can run a factory viscous limited-slip differential? While not perfect, that device will take some of the accelerative forces and transfer them to the underutilized inside rear tire.

So, we now have a car that should corner faster in steady state and accelerate earlier off the turn. Next installment we'll look at potential power outputs (complete with dyno graphs) to see who's faster in a straight line. Stay tuned... 

Power Analysis Jan 12, 2006

Last installment we established a competitive target weight for our project from a handling perspective. In that analysis we found that we could get our Miata to a legal weight of about 1950 lbs with some straightforward modifications. Today, we'll look at some power predictions to see if the drivetrain is up to snuff.

The STS2 engine allowances are fairly similar to those allowed for Spec Miata (SM). For each, the motor internals must remain stock, as per the factory service manual. Externally, each is allowed any cold air intake system which utilizes the factory airflow meter. SM is allowed any clutch and may remove the catalytic converter. STS2 requires those parts to be OE-equivalent, but instead allows for any exhaust header. STS2 also allows for any ECU programming within the OE box, while SM optimizes fuel and ignition through timing bumps and airflow meter calibration. Given the tradeoffs of each, power output should be relatively similar.

One of our project partners,Applied Racing Technology, provided us with a dyno graph of one of their typical SM motors. We then took a similar graph from our fully-prepped STS Civic and overlayed the torque curves. One caveat is that the SM test was done on a Dyno Dynamics eddy current dyno, while the STS session was from an inertial DynoJet. To normalize the curves, the Civic numbers were converted downward using typical conversion estimates. We hope to redo the test in the future on the same dyno once our STS2 Miata motor is complete.

Looking now at the graph, there are three mostly horizontal torque curves. One is a bone-stock Miata motor, the similar but higher one is the Spec Miata, and the one with the hump in the middle is the Civic (ignore the two HP curves). You can see that the SM makes similar peak torque to the Civic, but offers a much broader torque curve especially in the range from 5000-6000 rpm. It also does a bit better from 3200-4500.

One cause for concern is that the Miata drops off very quickly once it gets to 6700 rpm, while the Civic revs on nicely. ST Hondas typically have their rev limiters raised in the vicinity of 7800-8000 rpm to avoid costly shifts to 3rd gear (and back). If the Miata wheezes too hard, we may have to shift more which could be a hit against it. Our hope is that this can be helped out by installation of a good long-tube 4:1 header in the Miata since there is plenty of physical room for it prior to the cat. More on that in a future installment.

Now let's look at gear ratios. The Miata's first two gears are 3.136 and 1.888, with a final drive of 4.3. Multiplied out that gives 13.48 and 8.12. The Hondas have 3.250, 1.894 and 4.25, respectively giving 13.81 and 8.05. So in typical 2nd gear acceleration off the turn, its essentially a dead heat. The Hondas do enjoy an slight advantage off-the-line in 1st gear, but that is probably negated by the superior launch of the RWD Miata.

So, who's faster in a straight line? Since the CRX is potentially 100 lbs lighter than the Miata, and since they both make the same peak torque, one would think the CRX would be the winner, right? Well, not really. Its actually the area under the torque curves for each that is important. 100 lbs is about a 5% weight difference. And from 3200-5800 rpm, the Miata produces 3-6% more torque, excluding the Honda "hump" zone. So it makes up for lower weight with higher torque throughout much of the range. And given the level of accuracy of our paper analysis, its really too close to call.

So enough analysis, let's get to work! Next installment is a visit to the engine shop. 

Building The motor Jan 17, 2006

In previous installments we discussed our plan for the project and also our competitive performance analysis and goals. Now its time to jump into real race car building, and at the heart of any competitive race car is the engine. This is especially critical in low-power applications like STS2.

We dropped the motor off at Applied Racing Technology several weeks ago, and stopped in this week for a progress report and to snap some pics. ART is currently in the process of building five motors, mostly for road racing Spec Miatas, and we are in the middle of that queue. We had a good talk with ART co-founder Ed Gilfus about our motor and ART's philosophy on building fast Miatas.

When ART builds a Miata motor, they focus on three main areas for performance: head flow, compression, and bottom-end sealing. Even within the factory service specs there are areas for improvement in each category. For example, ART has done a bunch of flow bench work to come up with a specific valve job that flows the best, yet is still within factory tolerance numbers. The combustion chambers are also matched in volume by adjusting the sinking of the valves. And then there are the straightforward things like shaving the head to the minimum service spec to optimize compression. In addition, attention is paid to parts that just plain wear out. Valves are checked for length, square, radius and seats. Valve springs are all checked for rate and square. Cams are checked for lobe height and obvious signs of wear, and lifters are assured of proper operation. Any of these that is worn excessively is replaced. For a top-shelf motor, many of these items are replaced routinely for optimal performance and longevity. While this is way into the decreasing returns curve of performance for the dollar, we have chosen to go there to establish an upper limit of performance for the motor within the rules.

After cleaning and examining the head, we were fortunate to see very little casting flash in the ports and nice transitions to the machined valve pocket areas. Some of the other motors ART is currently building are not as nice in this area. Our head's gasket surface is also very flat. The bad news is that our head appears to have been serviced once before and not so carefully. The thickness is right at the service limit on one side of the head and about .003-.004 thicker on the other end. Our combustion chamber volumes match this trend with about a 1 cc spread from the thin side to the thick. Unfortunately, it is virtually impossible to fix this by shaving at an angle since there is no excess material on the one end. It would take an amazing machinist, and a lot of luck, to achieve this without going under the legal limit. We'll just have to live with it.

One note on cleaning: To remain legal, it is critical not to use any sort of abrasive cleaning process in the port areas or combustion chambers. Anything which alters the surface texture of those areas can be construed as a modification and is protestable. ART uses a combination of hot tanking and a very mild bead blast using materials that are softer than the aluminum alloy used in the head. This is sufficient to remove the carbon buildup without altering the head itself.

Since the STS2 rules allow it and we are going top-shelf with our build-up, we've chosen to go ahead and bore the block the allowed .010 first factory sizing. While this is typically unneccesary (and not legal in Spec Miata), we want to get perfect bores for optimal bottom-end sealing. When boring a block, ART uses a custom-made torque plate to simulate the stresses of the head and make certain that the bores are still round when the engine is bolted together. Into that block we'll be fitting Mazda factory overbore pistons that are parts-bin matched to within a gram of each other. Since ART builds so many motors, they order parts in large batches (see picture) and can select from within those batches to match up component weights. Nice.

We are fortunate in that our crankshaft is very straight. Ed says he sees a lot of Miata cranks that have excessive warpage and require reworking. Our rods are also perfect right from the factory, very closely matched in weight and length. Ever-the-perfectionist, though, Ed will be assembling the lightest rod to the heaviest piston to get all the assemblies perfectly balanced, yet still within the rules.

When it comes to clearances on the Miata, ART goes fairly tight. Since our autocross application does not see the heat expansion that endurance racing Spec Miatas do, we can improve our piston sealing without risking heat-related damage. Bearing clearances will also on the tight end of the range because of oiling concerns.

While we were there, we looked over ART's injector cleaning and flow-matching setup. Basically, you send them a set of injectors, they clean and flow test them, and match them up in groups of four from their stock. Once again, another advantage of using a shop that does some volume in your application.

And one final surprise was sitting on the floor by our motor. A nice low-mileage factory-optional viscous limited slip and the corresponding stub axles! Woo-hoo! The Spec Miata folks all typically run the more effective Mazdaspeed clutch-style limited slip but STS2 cars are limited to factory-only viscous units. This piece was pulled from a Spec Miata built recently by ART and thrown on the junk pile. Their junk is our treasure!

Next we catalogue our suspension starting point based on two year's worth of CSP development. 

Springs Jan 24, 2006

As our motor progresses at the machine shop, we'll take stock of our suspension setup. The next few installments will detail our baseline which is the result of several years of competition in the CSP category. As we do so, we'll discuss areas where we expect to make changes.

First up are springs. The primary job of springs in a production-based race car is to define the range of motion of the suspension. Stiffer springs allow for lower ride heights without bottoming. The first year we ran the car in CSP trim, we settled on rates of 550 lbs in the front and 300 lbs in the rear. With this, we could lower the car's cg quite significantly, lessening the weight transfer from inside tire to outside in a turn. This resulted in more efficient use of our tires and higher cornering forces.

Unfortunately, we also had some inconsistent handling. This was especially true on bumpier surfaces. We did a complete analysis of camber and toe change versus ride height by using an alignment rack and varying the ride height. The results showed severe bump steer and roll steer within the range of motion of our lowered ride height. The Miata has a wonderfully benign design when used within its normal range of motion, but becomes treacherous when lowered too far. Unfortunately, the SP (and ST) rules limit our ability to fix this through traditional geometry mods.

Our band-aid solution was to run much stiffer the second year (700/450) to reduce our range of motion, and thus limit the amount of toe change. The car became much more stable and predictable. We were also able to run a bit lower on ride height, which improved our cornering grip even more. This will be our starting point for STS2 but we have some concern that it may be too much for our street tires. The super-stiff sidewall of the Hoosiers used in SP allows the springs to do their job, but an ST tire has a much softer sidewall. The softer ST tire then becomes the primary spring force when the suspension springs have a higher rate than the tire's rate. This results in a pogo effect since the tire's spring force is not properly damped as is that of a suspension spring. We'll experiment with this.

Another point to mention on springs is weight savings. While long springs will work, you really only need springs that will just fully unload at max dynamic suspension droop, and will not coil bind at full compression. To that end, a 6" spring is long enough for a Miata. Its lighter than the 7" and 8" springs that many people run. Along the same lines, we run 2.25" diameter springs instead of the more common 2.5" just for the weight savings. A look at the online Eibach catalog will give a good idea of spring weights. Its not a lot, but every bit counts.

The picture below shows our complete coilover setup. The shock is a Koni 28-series monotube race shock which we will discuss in detail as a future topic. Note the use of later-model (99+) Miata upper spring perches which provide three important advantages over the original 90-97 setup: 1) allows an additional .75" of suspension travel in compression, 2) eliminates the undamped rubber in the original mount, and 3) saves weight. Shock mounting bushings were fabricated by sawing a pair of urethane sway bar bushings in half. The bottom washer is also from a 99+ Miata. This combination effectively reduces bushing compliance yet still provides for the angular motion needed to reduce sideloads on the shock shaft as the suspension moves. The bump stop is cut down to the size needed to just touch as the suspension contacts the unibody giving us maximum use of the designed suspension travel.

Next time we look at sway bars and alignment.  

Next: »