JohnK

Didn't even THINK it was anything else . "A dirty mind is a joy forever."

Yup, the sensors are indeed cheap - however, there's the matter of the code needed in order to use them to create useful data FROM the sensor. Think for a bit about how you'd develop a wheel spin controller and how you'd plug it into the ECU and what you'd do to the engine to stop it from happening in a controllable manner ... But the good news is that such algorithms are developed by folk with interests similar to ours and who are putting such things out there as Open Software so others can share in their discoveries - maybe even folk like NASA have such things out there which we can have access to. But nett is that all this cool technology is far from simple - getting all the bits needed to develop a working system and then getting the system tuned so that it helps you rather than helps send you off into the weeds.

I updated my diff ratios spreadsheet https://dl.dropboxusercontent.com/u/49212285/AsstdStuff/MPH-RPM05.xls and the webpage https://dl.dropboxusercontent.com/u/49212285/AsstdStuff/MPH-RPM06a.html data I have a sheet from a Subaru Parts desk giving ALL the pinion-crown combinations delivered on the different Subaru models (back when I thought buying a gear set was the way to change my ratio.) These are: Ratio crown/pinion teeth. 4.444 = 40/9 4.111 = 37/9 3.900 = 39/10* 3.700 = 37/10* 3.545 = 39/11 *for a 2006WRX AT is 3.900, MT is a 3.700, both are (viscous) LSD according to the 2006 manual at WRKINFO.com. N.B. the following: - the RPM-MPH XLS file has human-readable names for all the paramteters in the sheet so you can see how I'm doing the calculations. That also means it's easier to make changes, like for different wheel circumferences. Note that Tire Rack gives (sometimes) REVOLUTIONS PER MILE for the different sizes it sells. This value is a lot more accurate since the circumference calculated from the wheel/tire diameter don't take into account the distortion due to load. -Primary reduction is different for different year ranges: 1:1 for S2000 engines 2000 - 2003 1:203 for S2000 engines 2004-2005 Now, the gear set data I have has codes for the different models which would be useful if anyone can decipher the codes, which I assume is based on encoded model abbreviations known to some others, but not me. Here're a few of the lines. If this can be figured out, it'd be very useful for people trying to find a diff with the correct ration for their needs. Another question, is this diff unique to Subaru or is it fitted on any other models? Applicable model G/R (TS+OBK).MT 3.900 RS.AT 4.444 RS.MT + (WRX+TS+OBK).AT 4.111 MT.205 3.545 S.AT.251 4.444 S.MT.251 +W.AT +AT.205 4.111 W.MT.251 3.900 5MT.205 3.545 6MT 3.900 RS.5MT +(TS+OBK+WRX).AT 4.111

I agree but ... The only way you're going to realize benefit from being able to change the dampening characteristics is through software that has been optimized for whatever it is you're doing. Maybe someone out there knows the math+physics involved in optimizing the behavior of a vehicle as it goes around a corner, but TomF's description, and he's talking about a complete system that's been very highly integrated and developed and highly tailored, indicates that this remains a very complex problem. So bolting on a set of active suspension units - and the computer that controls them - is not likely to automatically deliver wonderful results. My experience tuning ECUs convinced me that most people taking this route end up making their engines work worse than than they did before, in spite of the obvious promise of the technology - and tuning an engine is a lot simpler than tuning a suspension. Hence my feeling is that the technology is taking us into a new place and all the hands-on, directly-involved fun is in the process of changing into something else. For an old git like me, it's probably time to be put out to pasture.

A short while ago I finished reading Michael Collins' book, Carrying the Fire - about his experiences in the NASA mission and being an astronaut. He was the low-profile crew member of the mission that had Aldren and Armstrong land on the moon. Collins manned the orbiter and collected the two and brought everyone back home - twiddling his thumbs (kind of) while the other two were on the lunar surface. Anyway he spoke of the same sort of duality as he went from being a fighter pilot then a test pilot and then a spacecraft pilot on the Gemini and Apollo missions - and questioned what sense it made for him to go from "flying" to sitting in a tin can pushing buttons on command to talk to the computer. TomF makes our point for us, namely that the driving experience has changed. Rather than depending on exquisitely developed sensitivities and being gifted with a superior nervous system, it's now about understanding the interface between you and the car - going fast efficiently is a matter of understanding the software. But rather than sitting stupidly in place at a display as things happen on a screen in front of you, you get to feel the g-forces and attitude of the car, and noise and smoke and fury as it happens. But the technology creates a new thing, and it's really different than the directly-connected experience of driving what most of us, 'specially older us's', call 'driving'. The world changes....

When I read this, because I'm really interested in things that have to do with a car's handling, it made me think of the different high-end models (Honda, Audi, Porsche, ...) that have electronically controlled limited slip differentials, and electronic control giving dynamic partitioning of torque between the front and rear wheels in four wheel drive cars. And then there are all the wealthy dentists, plastic surgeons, et ala. who, with several orders of magnitude difference between their bank accounts and their driving abilities, can safely drive about in sporty cars of all sorts that have well in excess of 500 horsepower without managing to become a grease-spot on the side of some building when they happen to tap the throttle and not be paying attention where they're aimed. Automobile had a good article this month on just that subject with a consideration of cars having too much power. Since I'm a scientist by training and disposition (and with my experience tuning engines with computer control) I have no doubt that this leveraging of technology by the automobile industry has just started and that the benefits will be amazing as things are developed for the masses. (DARPA has developed all sorts of incredible devices for NASA and Defence that are just waiting for some entrepreneur to turn into the next big thing - there's no way Jobs et ala. could have built the i-Phone without all the fundamental discoveries in physics and electronics and created mechanisms that our tax dollars financed, making the technology that, say, allowed us to put Curiosity on Mars so perfectly) . Cars will deliver unbelievable driveability, not only with engine responsiveness and power, but with incredible balancing of traction, control and providing superb driver feedback. But when I think of all the fun I had getting my Ducati running and working some 50 years ago, I'm not gonna miss living in a world where all the adventure I've come to love about working with cars and learning how to drive them will be taken away by engineers, computer programmers, and the effectiveness of our capitalistic economic system. Or more simply, Genug ist genug, zu fiel ist ungezund (Enough is enough, too much is unhealty). http://phys.org/news/2013-11-unconventional-car-transmission-differential.html

Oh boy - did I open a can of worms, now that I think about it. I had no other choice but to move my rack since the way the car showed up, the tie rods would have run in to the front link of the lower a-arm had I tried to drive around a corner ... Along with that, when I cut out the front end and made it like a standard S2K (if there is such a thing) I found that, no matter what I did, I recorded bad enough ride steer that I started exploring other solutions. The general specs on the Flaming River Cobra/MGB manual Rack & Pinion: 3 turns lock to lock; 6.25" travel; 3/4" 48 spline interface (Sweet carries such fittings). First of all, I would very highly recommend going to http://www.woodwardsteering.com/ and finding the section "Basic Rack and Pinion Tech" and reading "Configuring a rack for your race car." There are many many pieces on the site that give the best info on steering issues I've ever found - and that includes the works of Carroll Smith. One, "Steering ratios for oval track racing:", gives a superb explanation of the importance of quick steering. While all of this info concerns dirt/oval track race cars, the principles are perfectly applicable to sports cars. ...Steering systems are one one of those things that give new meaning to "everything is connected to everything else". The number of turns lock to lock depends on the pinion ratio of the rack (degrees of steering wheel movement to inches of rack travel), the steering ratio at the hub (the distance from the end of the tie rod connection to the hub to the kingpin line - or steering arm length), and where your steering stops are positioned, which determine your turning radius. There's a little trigonometry involved with the how much out of parallel the rack is with front axle line which shortens the effective rack travel, and this non-parallelness also influences how much Ackermann there is (my experience with this is that you want a lot of it). The rack travel has to be long enough to move your wheels to their steering stops. If the rack travel is longer than you need, your lock-to-lock value decreases. Generally, if both or youroutside front wheels turn to 20 degrees or so when at full lock and you've got less than three turns lock to lock, you've got pretty fast steering. Quick steering is usually very desirable (certainly in a Se7en-type car) but getting that is a lot more expensive than just buying a rack with a high pinion ratio and/or fitting a really short steering arm - the faster (and hence the more fun) the steering is, the more effort it takes to turn the wheel, 'cause you're moving all the pieces faster and generating greater turning force faster. Having to use more force makes you less accurate/precise in piloting your car - finesse and brute strength usually don't occur together. As you reduce your caster toward zero, you make it lots easier to turn the wheel (witness any of the videos of people tossing Caterhams about on wet roads) - but the caster has to be dialed in really really precisely and both sides have to be very close to the same, and this is a LOT of work - 'specially when changing caster will change your camber, and correcting the camber change will change your caster. You also lose directional stability and the car can become very twitchy feeling with low caster. A very effective solution is to fit power steering which takes the work away and makes low caster unnecessary. I chose hydraulic, but there are electric solutions too. Porsche likes electric, Ferrari likes hydraulic...) All of this gives you a great response to the guy who says "What kind of fool would put power steering on such a light car?" I wound up with 1 1/4 turns lock-to-lock, which is the same as the McLaren MP4-12C. Very surprisingly, and I've explored this on a bunch of different roads, I never got the feeling that it was too quick - after I threw away the tires that came with the car and which were dead by that time anyway. For those not easily bored I wrote up my little adventure in this area at http://usa7s.com/vb/showthread.php?t=7989&highlight=note+steering Anyway, carried away as usual, I hope that some of this was useful.

I just sold the rack and all related bits from my "kit" to a guy who's restoring an MG-B (and is putting in a V-8 that's known to fit ! ). To confirm, he brought the bent rack from his MG in for comparison and it was clearly a match (though the steering input shaft was real long on his). .........So what are you doing that you have your R&P out? I (in my usual gonzo developer approach) did my steering from scratch and found a company that makes really really fine racks if you want to engineer out all bump steer and speed up your steering (but they're a bit more than $129)- it's out there written up on this site.

Arithmetic: the spreadsheets I posted have the arithmetic I used in the cells, so if you download the *.xls file you can use it as a reference. A while back Loren saw the figures I posted for this thread and said they were accurate, given slight variations for the 17" tires different people run. The Primary Reduction comes from a third shaft in the trans. The Mainshaft plugs into the clutch, like god intended, and the Mainshaft drives the Countershaft. The final output is not the Countershaft but a third shaft that's driven by the Countershaft, and that's where that additional reduction comes in. A while back I accurately measured the different ratios on a Ford Type 9 (Superformance S1) by attaching a degree wheel to an axle and then turning the input to the diff exactly one full turn, if I remember correctly, and then reading the degrees off the axle. Lock the axle you're not reading from to prevent its movement. Reading a manual I got from WRXINFO website (earlier) a 2006 Impreza WRX with automatic trans has an LSD (viscous) and a 39/10 (3.9000) ratio.

Thx, but if I were GOOD it wouldn't be taking me so long. You should have seen the stumbling around I did for months and months just trying to figure out how things worked and trying to understand what Carroll Smith was laying out in his "--- To Win" books (which I very highly recommend). If you admire what I've been able to put together (an old, grumpy, retired Biologist) the lesson you should take home is, "Hey, if THAT guy can do it, I should be able to too!" There's nothing quite like struggling with a problem, figureing it out and then taking the car, Your Car, out and having it work like you want it to.

Couple of thoughts, pretty general but perhaps some helpful bits. I blended a deflector+screen into the rear edge of the hood of my WCM S2K by attaching an aluminum frame to the fiberglass of the hood with 3M structural epoxy*, and the frame was fabricated with tabs to connect to polycarbonate screens (with scratch resistant layer on both sides **). In my particular case, I think the saving virtue (and also the thing that made it really challenging to build) was the curve at the rear of the hood that I fitted the aluminum frame, and then screen, to. What happened was that TWO curved surfaces were joined when all was done. This made the thing into a structure which has proved pretty rigid (no flutter at 100+ MPH), and it directs the air around the cockpit pretty well (but not perfectly, 'specially at speed). Polycarbonate is very strong, and being curved adds a lot to its rigidity. Shots at http://usa7s.com/vb/showthread.php?t=7421&highlight=Ithaca. *3M Scotch-Weld Epoxy Adhesive Tube Kit. 2216 B/A gray - Gives 90 minutes working time - and 7 DAYS to cure (but you can speed that up a lot with moderate heating). **This was a Bayer product the data on which I've misplaced. You can get info from my local source at Queen City Polymer 513-779 0990, but I'll bet any industrial plastics source can be a provider. 1/8 x 24 x 48 piece (enough for 4 screens for my build) was $71. Secondly, if you decide to get into fiberglass to make a diverter, try http://www.fiberglast.com for an overview of what you can attempt, and for more modest work ITWEvercoat (formerly Fiberglass Evercoat) has perfectly good resin products that you can get locally anywhere. I made a NACA duct out of fiberglass to provide an effective (and cool-looking) means of getting under-hood heat out of my engine bay. Using fiberglast's info (they have how-to videos that are really informative) and products I was able to do the duct via something called "vacuum bagging" which allows you to make some pretty professional, not to mention very functional, pieces out of resin and fabric. Pretty much of a job, but for a one-off project and modest expense it was a good experiment - being successful the very first try. Best of luck.

These data are accurate, WestTexasS2K confirmed them a while ago. https://dl.dropboxusercontent.com/u/49212285/AsstdStuff/MPH-RPM02.html If you want to change values for things like your tire's fitted circumference (and you calculate from ROLLING radius, which is not equal to nominal radius) and this works in your spreadsheet try https://dl.dropboxusercontent.com/u/49212285/AsstdStuff/MPH-RPM02.xls or https://dl.dropboxusercontent.com/u/49212285/AsstdStuff/MPH-RPM02.ods Tire Rack, for some of their tires, gives tire revolutions per mile - as an example of the error involved between nominal radius and rolling-radius based circumference, the difference between the published diameter of my Yokohama S.drive 245/45R17 (25.6") and the "revolutions per mile (812) comes out to about 3%. Yeah, I'm being a nerd here, but little things can add up and as the good Mr. Smith continually tells us, "Races are won by tenths."

I was looking at this thinking, "Man you just never see stuff like that in Cincinnati." And, there really aren't any car places that I know of here that have things like this sitting around for folk to come look at. But then again, within walking distance of my house (which is pretty modest all in all) are a Porsche and Maserati dealership, and a couple of blocks further down is Zakiras Garage which has done restorations for the Collier Museum in Fla. and at any given time is like visiting behind the scenes at the Smithsonian. A few years ago the Cincinnati Art Museum had a Ferrari P4 (from the days before aerodynamics made their influence on design) on display for a while, borrowed from a local collector just up the hill from me - which I went drool over, well, I don't think I actually drooled, but it took me days after the visit to get my concentration back to other stuff. I think, because of the conservative influence of its German heritage, this city just isn't into to showing off. An early jazz radio station here used to announce their call letters by saying "This is 740 AM Stereo WNOP, broadcasting from behind the sauerkraut curtain".

Steve, you've got a good eye for picking out very fine looking cars.

Looks great and it completes the rear part of the upright very nicely. Well done! But on top of that, think about: when anyone looks at the bits that make up the workings of current cars, especially sports cars, it's clear that you'd need not only a $5K CAD package to draw any of them but a Finite Element Analysis package to tell you where to make it thick or thin, not to mention an engineer or two to give you the loads that are going to be fed into the part, and of course an NC 6-axis mill to turn it out ... all of which is way beyond the means of anyone whose name isn't Boeing or General Electric - or even McLaren or Mosler. What you've done is the kind of stuff that's in the realm of possibility for an enthusiast with his head on straight and a desire to have his car work like he wants it to. And what that's doing is recreating the good old days when you could make something in your back yard with your own two hands and some affordable tools that fixes something properly. And I happen to also think that that's great fun!

How'd you find that an '07 Forester has a 4.11 + LSD? My take on the oil business is that light viscosity (relatively) and good lubrication buy you really tiny increases in power, but unless your competition is really wheel-to-wheel competitive, you're not going to see any difference. Major brand is just fine unless you're just after bragging rights (which are cool also :-) ). Subaru says 0.9 U.S quarts. But, if you're going from a 3.54 to a 4.11, you may think you've died and gone to heaven re the difference in acceleration. I'm figuring to make the same change if my time doesn't run out. Kitcat's post-drive comment about my performance being in the "wuss" category pushed me to it.

When I was redesigning everything in my footwell I found I could put a small tab above, to the left and a bit rearward of my clutch pedal that I could catch with the toe of my shoe. Since I was fabricating everything else in the area, doing a rest was relatively easy, 'cept for the time spent figuring out where the dickens I could position the tab so that it would NEVER be in the way if I was in a hurry to do something with the pedals, but still allow me to actually rest my left foot on it when cruising. It turned out to be worth the effort (and I seem to have gotten quite used to being able to do reasonably solid welds while cramped up into one weird position or the other :-) ). A note on room for your feet: Kitcat in first introduction to my car said that the footwell in my S2k was really cramped compared to his his std Caterham ( keep forgetting the type but it's NOT the one with the roomier chassis). So I guess you can find space for such things if you really work at it. Now a real dead pedal is another matter entirely.

The rack is supposedly a copy of a stock MGB. If you have an imports shop around you should be able to get them there. I priced a pair for about $35 a few years ago.

Break my heart! My first car was a 1957 MGA, back in 1962. I was the 5th owner of it, if that tells you anything about the shape it was in. I STILL remember the first time I drove it as clearly as if it were yesterday. The feedback through the steering was so fine that I felt like I actually had my hands on the road, and is the reason I bought my Se7en, and spent so much time getting it to handle and feel like I remember the "A" felt (and it damn near does!). Made me the driving fanatic I remain today . Was visiting our local "automotive shop" treasure, Zakiras Garage, a while back. They were restoring a 1934 MGP and which raced in a Le Mans a few years later, for the Collier museum - and noticed that it had two 6 volt lead-acid batteries running in series, one of each tucked behind the driver's and passenger's seats. So my "A", and I assume yours also, has deep roots. Congratulations, and many happy miles.

Perhaps there's a difference in parts here (there's a mechanical and a hydraulic LSD for instance), however....If everything is intact and connected, where can the axle move to? The diff is fixed in place, and, at the axle's other end the upright is located laterally by the lateral links, the radius rod (runs fore and aft) keeps the upright from twisting (more than a small amount), and the shock limits the vertical travel of the upright. If one were to plot a view of the axle pivoting about the diff through a bump-droop of +/- 3" I don't think you'll see a change in the hypotenuse that's enough to equal a distance sufficient to get the axle free of the diff. However I look at the assembly, I keep coming to the fact that the axle is trapped - unless there's too much room for motion in plunge, which indicates that something's moved. Again, possible part differences, but both of my axles have circlips, and every time I pulled either, it took no more than a modest levering with a large screwdriver to pop the axle out. I've never seen an assembly that was retained in place against any kind of operating force with so insubstantial a restraining element. Actually, with regrets that you're having problems, the prospect that something else is happening is intriguing. If I'm mistaken, it'll be really interesting to learn of another property of this system.

Roman, When I first got my car I spent more time than I care to think about on the suspension - a bunch of which involved taking fairly accurate measurements of everything in 3 dimensions (Took three attempts before getting to where I trusted my measurements and understood the limits of the accuracy I could deliver) so I could enter reasonably accurate data into my copy of WinGeo3 to understand what the suspension was doing, as well as understand the fundamentals. While this was a ton of work, it's paid off over the years and has proved its worth. In my experience, you just gotta know what it is you're working with, otherwise you're just throwing mud at the wall. "The circlip" is supposed to be on the inboard end of the axle (see below). Normal behaviour is that the axle will resist coming out unless you lever it, just a bit, with a pry bar. When you work with the rear a bit you'll see that the circlip is just a convenience for assembly since it holds the axle in position while you're working on things. The resistance you encounter when removing it lets you know that the axle was seated and the same when you install it back into the diff. My Civic is the same, and I expect this is a common design. What keeps things together is the suspension, and its alignment. When I changed one of the axles on my Civic I nightmare'd about it coming out before I realised that it couldn't - the suspension links kept it from doing so. The same SHOULD be true on our cars, provided everything is aligned correctly. However, these cars have so much adjustment available that you can put things so far out that they might as well be on the Moon. The measuring I did early on allowed me to align the car "perfectly" - when I finally took it to a local shop with the latest Hunter equipment, it came out on the money - shop owner said he couldn't have done it more precisely although "accuracy" is another matter since it's determined by your camber curves, tires, ..... . Getting to that point involved replacing several suspension links. My measurements, mapped onto the floor of my garage allowed me to draw a rectangle that was true to the long axis of the car and which I used to put each wheel exactly (more or less ) at the corner of a rectangle that gave the true long and lateral axes of the car. For my car, to position a wheel pretty exactly at each corner took cutting one of the fwd lower rear suspension links to add 3/4 of an inch in length, and replacing a number of the rear links. I found that, if some of the as-delivered rear links were used, not only was the wheel too far inward re the reference rectangle, when correct camber was set the axle was pressed into the diff causing bind in its travel (and I assume bearing destruction eventually). In the other direction I found (stumbling around more than a little) that what looked like a good adjustment was acting to pull an axle out of the diff. Since we use standard (we hope) axles, the position of the rear wheels is determined by the location of the diff and the lengths of the axles. The links have to be the correct lengths for the wheels to be in the correct positions. "Correct position" is not only at the right width, but also fore and aft - it's real easy to have the rears parallel to each other but off center with respect to the long axis of the car, which gives you weird behaviour coming out of corners and such. So, based on my experience and probably obviously true to any mechanical engineer who gave half a thought to it, the only reason axles "pop" out is because the suspension is set incorrectly, or something fundamentally is wrong with the suspension. When the suspension goes through its travel, the axle moves in and out of the diff - hence the length of the splines. This is called 'plunge' and my suspension software calculates a value for it. The problem is that there's not a gauge that I know of that'll tell you where in the range of plunge you are when you align your car. As a general rule, whenever I change anything on the suspension, after it's buttoned down, I grab each axle and push and pull it to make sure there's a little free play. If it's binding, there won't be any movement. I think that the worst (bind) case is when the axle is level with the diff. If there's only a small amount of play, I reason that it's improbably the axle could un-plunge enough to get out of the diff. Along with this I've obsessed about suspension travel, not wanting, ever, to allow things to go solid - silasto deals with full bump wonderfully and the shock maker handles droop. I had my shocks built to allow only 1 1/2" droop because of the ride steer problem in droop with our/my setup. These circlips are a pain. Since they don't really do anything when the car's working (i.e. they don't bear any load) they're not tempered and when an amateur mechanic like me takes one out, it's soft enough that it can't be re-bent properly to seat in its groove in the axle and may go in hard enough to make you worry, or get knocked out (possibly, but maybe not). But they are supposed to be there, the Subaru manual says replace when removed, and they guide assembly for the mechanic. And, they may produce shiny bits when digested by the differential (sorry about that).

Hi Jim, Glad you found my original post interesting and thank you for the compliments. Original plan was to buy a "distressed" Superformance S1, make it 'right' and then experiment with the suspension. The Ultralite made too much sense re its basic features (engine I would have no reason to fiddle with, IRS, LSD, and it looked OK - however ....) Everything I've tackled has been out of necessity (several were no-brainers), but everything was done after careful study, consulting what's considered 'good practice', and my conclusions and intent corroborated by my Mechanical Engineer friend, - but I will accept that I used each fix as an opportunity to spend time learning about what constitutes a good design, and to develop new skills. (I mean, I'm a Biologist by training and my accomplishments have to do with experimental surgery and analysing nervous system function!) A thought or two re your bolt breaking: With Loren's experience that he's never seen this - and he has many instances where the rear suspension worked under very high loads, and your observations that the suspension link in question and the rest of the linkages look as though they act as intended (and the link in question is loaded in compression and tension only, in its travel), there might be something else taking place. A while back there were a couple of posts re an inboard suspension link being pulled out of the chassis (broke the tube on either side of the clevis), and I thought about the shock/a-arm going solid at the limit of the suspension's travel as a culprit. Re the rear, I saw that as I aligned the suspension trying to get all 4 wheels to the corners of the car, it wasn't hard to move the upright inward enough that the driveshaft plunge was too much and the D/S would impede the movement of the suspension. At the time my immediate worry was about ruining the CV joints and/or diff bearings, but such would also load the suspension in ways that were likely self-destructive. Also, as I got Koni to make me a set of shocks and had some heavy-duty software to inform my spring and Silasto choices, I made sure that the shock handled all at-limit suspension travel, avoiding any 'going solid' conditions. Hope this helps inform your thinking about the failure.

Can't tell from the photo, but it looks like the welding was done on the shaft of the bolt, since I can't see the head of the bolt. All of mine were done on the periphery of the head - which gives a much broader footprint for the weld on the clevis. Additionally, I can't see that there's any penetration on the clevis. The welder should have melted metal on both parts and the failure would have torn metal from the clevis. Speedway. Weld-on shock mount with spacer. Part number 917-21000 $5.99. Suspension links (control arms) part number 910-34158-(length in whole inches)-ALU. 6061-T6 aluminum $9-12 depending on length.

Recommend http://www.airpartsinc.com . they carry a wide range of Normalized (heat-treated) 4130 steel and their clients build aircraft. They give good service and are nice people.

Late to the party, but... Go to my page http://usa7s.com\vb\showthread.php?t=7421 ; then find the link just above the first pic ( and following "More pics at ..."). This'll take you to a page of text with links imbedded. At the end of the text look up a bit and you'll see a link starting with "t50..." and describing "single shear" - which is the problem here, and its solution in this situation. The link starting "t55_a..." shows a pic of the solution. Further up in the text is a top view at "t50..." Boot machinery.