Here I will document the progress made on the rear strut bar. Everything will be listed in reverse order so the newest news is always at the top.
Well it was a beautiful morning today. You know, that first one of the season that you've been waiting for for weeks. For me, this was it. So, I went down and started setting up the Paseo for testing. The setup went well and easy enough (see picture below). I ended up using a very small copper wire for the 'string' to eliminate any stretching inaccuracy. So, I got it all setup and took it out to an abandoned parking lot and started driving in circles. What I did was speed up and start turning in a circle. I then accelerated until I started to loose grip with the pavement. At that point I've already passed the threshold of maximum traction and thus slowed to a stop. I should mention that this test was done on my 1997 Paseo. As far as suspension modifications it has TRD springs and KYB GR-2 struts. The tires are all season Kumho ECSTA HP4s on stock Paseo alloy wheels. The results? Well, take a look for yourself.
Video one is the string test. Unfortunately it is somewhat inconclusive as there is a lot of vibration. It is hard to tell very clearly if anything is moving and if it is how far. However, as I said before, if the towers moved even 1/16 of an inch the string would drop approximately an inch. I think its very easy to see the string did not move nearly that much.
Video two is the dial indicator test. This ended up being quite accurate and the more conclusive of the two tests. The first big hash mark on the indicator is a movement of .010". Thats about as far as it goes. What does this mean? It means the strut tower was bending over at just under 1 degree! At the height of the tower 1 degree is about .009" of horizontal movement. Thats it, thats all the movement there is. Its pretty much nothing at all.
So, with this new evidence I see almost no reason to install a rear strut bar. Only in the most extreme situations would it be useful such as an all out race car on a road track. It may be useful for such a car as they will have a much stiffer suspension and much grippier tires and thus more stress would be put through the strut towers. However, for anyone that is doing less than that I see no reason to install a rear strut bar. And, with that, I wrap up this project.
I've figured out how I am going to be testing to see if the strut bar is needed or not. I will be doing two different tests to try to measure how much bending is going on in the chassis. They may seem simple and even a bit cheap, but it will be effective.
The first method is to simply tie a string to both sides of the rear strut mount bolts. In the middle of the chassis a ruler will be mounted vertically to measure the movement of the string. If the string droops under hard handling the strut towers are obviously moving. To what extent has yet to be found out. To add accuracy to this test I'll probably use a rubber band to pull the string down and keep it taunt under all situations otherwise the readings may be inaccurate with the harsh side to side motion and slack in the string. This string represents the exact forces that the strut bar would be absorbing and transferring to the other rear strut so it is a very good test. It should also be quite accurate in measuring movement between the struts. Any small movement will be very visible on the rules. Here is are pictures to illistrate what I'm talking about. The two 18" dimensions represent the string being 36" long. At a 36" distance (3rd picture) the string would be taunt and straight across. At a 35.94" distance (4th picture) the string droops approximately 1" down as shown.
The second test I'll be doing is a little more sophisticated and accurate, however it may not be quite as applicable as the string test. This will also measure if the strut is moving. the plan is to use a magnetically mounted dial indicator (device for very fine measurements down to .001") and mount it on the side of the strut. The indicator will be butted up against the floor of the chassis and will measure if the strut tower is tipping at all. Any movement will be seen on this indicator. As I said before, it can detect even .001" of movement.
Time for a handling modification. The point of this project is to make a strut bar to stiffen up the rear end without having to cut anything out of your car. I've actually been thinking and working on this for a while now. There have already been a few designs that I have gone through that I thought might work. Each time I scrap one out for something I think will work better. Thats how it goes.
There is actually already a bar that appears that it would work on a Tercel or Paseo without any cutting modificaitons. It is made by the well known brand Cusco. However, if you know a few things about material strength and stress you can see that this bar is far from the ideal design. The bending stresses that are induced via the 90 degree arm near the mounts is not a good design and it will not hold up to stress nearly as well as something without that kind of arm on it. There are a few other rear strut bars made for the Toyota Starlet, but they require cutting into the rear frame so the bar runs directly from one strut mount to the other without any offset. As mentioned before I am trying to avoid this, so these bars won't work either.
So, it appears we need an alternate solution. That is really what this project is about. Of course, before I go forward with this modification there is always testing to be done. If testing proves that there is no need for a rear strut bar there is obviously no need to go any further with this. This is a much debated subject when it comes up in forums. Is the rear dash sturdy enough to keep both rear wheels firmly planted or not? Unfortunately, no testing has ever been reported on this subject to prove one way or the other. This brings up the first phase of the project (as it should be in any project), testing. To do the testing I will need some way of measuring how much the rear strut mounts actually move under hard turning. However, I'm still figuring out how this will be done accurately. If you have any suggestions feel free to shoot me an email at firstname.lastname@example.org.