[Tycho]

For background information, start here:
http://www.insightcentral.net/forum...lternator-motec-management-33.html#post646729

and here:
http://www.insightcentral.net/forum...ront-rear-springs-new-rear-anti-roll-bar.html

One typo in Julian's dedicated spring post is that the rears are +55% rate and the fronts are +45% rate (he wrote it the other way around, but if you work out the rates from the first thread, that is what they work out to). But basically +50% each.

Went junkyard hunting today, looking for a nice front spring replacement to go with the Matiz rear springs I got from Scott.

Looking for a 1" front lift and slightly higher rate (I personally am not looking for +50% rate per-se, as the rear springs are).

I am also looking for a rear bar of sorts. I have previously played with swapping rates and bars on a couple Saturn S cars (incl a nifty 4 way adjustable bar, which was handy when the seasons changed to ice and snow) and some older Subaru GLs, which had tons of different spring and swaybar options for the coupe/sedan/wagon that all shared suspension parts but all had different rates and thicknesses.

I took an attempt at some very ghetto spring rate measurement, but looking at the data in excel tonight, I can say it is complete garbage. I'll come up with something better.

FRONT SPRINGS

Stock springs:
Insight Front
Wire diameter: 10.5 mm
Free length: 282.5 mm
Spring rate: 112 lb/in (per Julian's measurements)

Spring options:
1. 2002-2008 Corolla CE/LE/S (the crappy ones) rear spring
Wire diameter: 11 mm
Free length: 335 mm
Spring rate: 162 lb/in (according to Julian)
Very close to measurements:
MOOG® 80667 - Toyota Corolla 2005-2006 Rear Coil Spring Set

These guys just fit over the Honda strut (on the lower end) and need to be trimmed and flattened on the top to fit against the strut top (as per Julian's instructions). They are slightly smaller in diameter than the unusually small (for a front spring) Honda OE springs, and may rub the plastic bellow just a tad. They actually do not "seat" all the way down on the strut with the plastic bellow in place, but that could be resolved with a little trimming. An interesting thing to note is that when cutting 1.5 coils off, the spring rate is going to increase over what it is already, as the total length of the wire is going to be shortened (you can think of a coil spring as a straight wire "spring" that has been wrapped up for packaging reasons).

2. Toyota Echo rear springs
Wire diameter: 11.8 mm
Free length: 322.5 mm
Spring rate: 117 lb/in (according to some people on Yaris forums)

These guys are basically the same diameter as the factory springs, but have tight coils on either end that would need to be trimmed down. One coil would also need to be flattened as with the Corolla spring. In the process of shortening this spring to the right length, I think it would gain maybe 10% in rate (as you'd be taking 10% of the coil length off). This would make for a 130 lb/in spring which would be a nice amount over stock, perhaps without being too much for the OE dampers.

Other promising looking options I was not able to grab (ran out of time):

Toyota Yaris rear springs. These measure a tad thicker than the Echo springs, at 12.3 mm diameter. Other dimensions looked similar.

Toyota Matrix rear springs. Identical to the Camry ones, but 11.9 mm wire instead of 11. Even higher rate basically. The free length may be longer as well, but probably close.

2001 Civic (and other Civics of this generation). These measured at 12.2 mm diameter wire and other dimensions looked like they would allow for a straight swap onto the Insight front strut. The overall diameter of the coils was smaller (about 4.5" vs a hair over 5" stock) but the spring seat areas looked dimensionally very close. Top of these springs looked to be pre-flattened. Don't know their actual rates and free lengths are tho, which are key.

Mazda 3 rear springs. These also look like a dimensionally close fit, but perhaps too small. Wire diameter is 11.15 mm.

REAR ANTI ROLL BARS

The following dimensions should be adhered to when making a rear anti-roll bar for the G1 Insight:
55 mm rod end "dogbone" height (this is the height inside the rear torsion beam)
800 mm total length (center to center of the mounting holes)
Toyota anti roll bars use 13 mm shank diameter bolts (and something like 16 mm diameter tubes on the dogbone portion). Would be a reasonable setup if someone had an Insight bar fabbed. The Toyota bolts should be readily available.

The Insight beam has some braces near the ends that limit rod length to 800 mm. These braces are tapered pieces of steel welded into the beam. They start 350 mm from the center of the bar in each direction (so they are 700 mm apart where they start). They taper rearward inside the bar, so if your anti-roll bar is too wide, you will not be able to tuck the ends "inside" of the "C" portion of the torsion beam, as it turns into an "E" at about 900 mm (as these braces taper thicker and eventually tie the torsion bar to the hub). See pictures below.

I snagged two bars from the yard today. They spec out as follows:

1. 2002-2008 Corolla CE/LE/S (the crappy ones) rear sway bar (and Matrix, same sway bar)
20 mm diameter
50 mm tall rod end "dogbones"
1020 mm total length (center to center)

2. 1st Gen Prius rear sway bar (yes, it had one, and the Echo didn't for some reason)
17.5 mm diameter
50 mm tall rod end "dogbones"
900 mm total length (center to center)

Either of these bars, with a little cut and weld, will work fine. They will need a thick washer on either "dogbone" end when slid into the beam, but no big deal. I am planning on having both modified to fit so I can try them both out and decide what degree of lateral coupling I like better (how much load transfer I like).

There are a couple bits of information I don't have that would be terrific when making my spring cutting calculations and measurements. First, what is the length of a stock Insight front spring when the car is sitting on level pavement. I didn't bother to take this today, but if it's dry out tomorrow, I'll probably get it then. Second, what is the weight distribution of the car. How much weight is on the front wheels? If I have this information handy, I can calculate how long to cut the springs in order to give me the correct installed ride height (I can back out how much the stock springs are compressing, or back out how much load they are carrying if I know either the load or the length).

Pics below are:
1. Today's Haul
2. Insight strut taken apart (showing the bumper the car basically rides on)
3. 1G Prius anti roll bar diameter
4. 2002-2008 Camry/Matrix/Vibe anti roll bar diameter
5. Inside one "end" of the 1G Insight rear torsion beam (showing internal flange that limits length of rear anti roll bar)

 

[Tycho]

A couple pics of the Corolla spring going on the Insight strut here.
1. The spring going on bare (note how it fits all the way into the spring seat).
2. The spring going on with the boot in place (it won't sit all the way down, some little segments of the boot could be trimmed so it will squish into place nice and snug).

 

[JulianEdgar]

I deleted the front bellows. You can get smaller diameter aftermarket ones that will fit.

 

[retepsnikrep]

Remember the MT & CVT front springs are different rates as the CVT weighs an extra 80kg all at the front.

So putting CVT springs on a MT would give you a bit more clearance/height at the front I presume.

 

[Tycho]

The CVT springs came to mind. Do you know if they are higher rate or just longer free length (Subaru wagon springs were done this way)? I'll work on sourcing some and measuring everything.

Going to rig up a proper way of testing spring rates here too.

 

[JulianEdgar]

To test spring rates I used my vertical mill and a set of 120kg bathroom scales. Beware: you need at least 50mm of deflection before you start to get anywhere near a linear rate reading.

Going up by 50 percent (ish) in rate front and rear sounds a lot but, on my car at least, it works incredibly well.

 

[Tycho]

An update outlining my current state of things and game plan.

After knocking the powder coat and rust off the bars, the Prius bar is 17.0 mm solid and the Corolla bar is 20 (19.9) mm solid. A solid 20 mm bar is probably of similar stiffness to the 21.5 hollow bar Julian found on the Australian Corolla.

I have had both shortened to 80 cm and am giving them a coat of rust preventative paint now.

Front springs length of stock springs with the car sitting on a level surface is 6.75" from the bottom of the lower perch to the top of the upper perch. The perches are each a about 1/4" thick (the top is more, the bottom is less), so the compressed spring length is 6.25" or so (maybe a bit less, but more than 6.0"). When the struts are unloaded this length is something like 9.5". With the spring unloaded out of the strut, free length is something like 11.125".

So:
Free length: 11.125"
Unloaded strut spring length: 9.5"
Loaded strut spring length: 6.25"

At 112 lb/in, we can infer something like the following:
11.125-6.25 = 4.875"
4.875" * 112 lb/in = 546 lb -> ~ 550 lbs -> 1100 lbs on the front tires assuming about a 1:1 ratio on the springs, which is close for the mac strut setup. Assuming 1850 lbs this would be a 1100F / 750R weight distribution which is about 60/40, which is close-ish.

Now, we can calculate (or, at this point, roughly estimate) the free length we want the Corolla springs to be assuming they are in the 165 lb/in rate, bearing in mind that they will actually be a bit higher as we will be cutting a coil or two off.

Working the numbers backwards...
550 lb of spring loading with a 1" lift means ... 550 lbs on a spring that is now 7.25" long (vs 6.25 stock). If this spring has a rate of 165 lb/in we can determine what free length we want it to be.

550 lbs * 1 in/165 lbs = 3.33" of spring compression
7.25" + 3.33" = 10.58" -> ~10.5" (remember, the rate will be higher because we are cutting it down some, so we'll low-ball the length by a hair).

The Corolla springs are 13.25" free length, so we want to cut 2.75" off ... or about 3" as the rate will increase by whatever we cut off over the total length ... so 3/13.25 = 23% stiffer ... or right around 200 lb/in when it's all said and done.

We could quickly run the numbers again for a 200 lb spring...

550 lbs * 1 in/200 lbs = 2.75" of spring compression
7.25" + 2.75" = 10.0" free length on the 'rolla springs.

If you do what Julian did, and cut 1.5 coils off and grind the end of the wire flat (instead of leaving it round), you get a spring that right at 11.5" free length. If you then do as Julian did and heat the end of the spring, then flatten it down against the the next coil (making the upper coil totally "flat" like the OE Honda spring), you end up with a spring that is just about dead nuts on 10.0" free length, which would give 1.0" of lift at 200 lbs/in with 550 lbs of weight on each front wheel.

Perfect.

 

[JulianEdgar]

Sounds good.

 

[Tycho]

Springs are now prepped, but didn't have time to get them on the car today.

I left them 1/2" long (10.5") and opted not to bend the coil flat as I did not want to torch the springs. Instead I ground the already flattened-ish side of the spring (the side you don't need to cut any coils off of) to be flat. I then tapered the other side of the spring (the non flattened side that sits on the lower spring perch).

The recipie is:
1. Cut appx 2.25 coils off the bottom of the Corolla spring
2. Grind a taper on the end so it seats nicely on the lower spring perch of the Insight strut
3. Grind the top coils flat so that they have more surface area against the rubber upper spring hat.
4. Grind the bottom of the end of the top coil so it does not come in contact with the second coil as the spring compresses. You don't want the last 1/2 coil on the top to become a "dead coil" as soon as the strut is loaded.
5. Paint

You should now have a 10.5" free length spring that will droop the first 0.5" pretty easily then be nearly linear at about 200 lbs/in after that.

Pics below. Install tomorrow hopefully.

 

[Tycho]

Springs are now on the car. Here are some pics of the modded springs and swaybars after a couple coats of paint, and a shot of one side installed on the car.

I'd note that this setup has very good coil engagement on the lower spring perch, even tho the spring is shorter than stock (free length is shorter) and the bottom coil taper is a little steeper than stock. I was a little concerned that the spring would "sit up" on the perch at full droop (as many ghetto fabulous lowering springs do) which can result in the spring slipping off the perch, but seeing it all assembled there is zero chance of that. There's probably 1" to 1.25" of compression on the spring at full extension of the strut, so it has about 200 lbs of preload on it and this is enough to really secure the lower coil almost all the way around the lower perch.

 

[Tycho]

Here's a couple shots of the front bump stops. They are progressive in three stages. The bottom stage is almost completely hollow. The next stage up is about 50% hollow. The top stage is solid.

Note how the bottom stage is cracked. The car essentially rides on that stage and every time you hit a bump it gets completely compressed and the next two stages are engaged.

I trimmed the bottom stage off to give me more wheel travel with linear spring rate (i.e. spring rate from the spring alone). Since the new springs are approximately 200 lb/in (vs 112 or so stock), a bottom out bumper that has an initial spring rate that is about 50% higher than stock will still be a "smoother" transition than the original setup, as my initial spring rate slope is nearly twice the original slope. Where the spring curves intersect (which happens when you engage the bump stops) the change in slope will still be less sudden than it was before.

 

[Tycho]

Finally, here are pics of the change in ride height. Using very crude calculations a few posts ago, I estimated that a 10" spring would yield a rate of 200 lb/in and a lift of 1.0".

I ended up cutting these springs down to something like 10.75", then grinding the already flattened (wound horizontal to the ground) coils flat (cross sectionally flat) and grinding the cut coil (the one that sticks out a bit) at a taper so it would engage the lower spring seat of the strut a little nicer. This resulted in a 10.5" spring when measured from end to end. Given how the coil sits in the strut and the fact that the top coil is a "dead" coil, I expected this to yield very close to the results I had calculated for the 10" spring (in other words, I was expecting the first 0.5" to be pretty soft, then linear after that).

The end result was about a 1.125" lift.

Here are the numbers (as measured to the little "point" from the front bumper in the fender well):

Right Side Stock: 22 3/16"
Left Side Stock: 22 1/16"
Average Stock: 22 1/8"

Right Side Corolla 10.5" spring: 23 3/16"
Left Side Corolla 10.5" spring: 23 5/16"
Average Corolla 10.5" spring: 23 1/4"

Net lift: 23.25 - 22.125 = 1.125"

Pics of my measurements below

 

[Tycho]

Notes on how the car feels with the new springs:

Julian definitely nailed it when he said it feels more "busy". It is surprisingly not too underdamped, which is what I was expecting, probably as the stock dampers are really fighting the rebound of the high rate bump stops most of the time as the car rarely makes it "up" onto the springs alone. Now that it is actually off the stops, even springs that are nearly twice as stiff have less peak spring force.

The "busy" feeling is mostly just the front end of the car actually being suspended. It feels like a small, light, nimble car, instead of an oversized, un-suspended go kart. It "settles into" corners now, instead of riding up on the outer bump stop, and the ABS doesn't engage when regen braking over minor road imperfections, as the wheels actually stay in contact with the pavement now. What a concept.

It feels like front end chassis roll is slightly increased, probably as the car rolls onto a spring, not a bumper, but the steering feel during transitions into and out of chassis roll situations is much improved (which is to say, it is now present).

I must say, HUGE thanks to Julian for playing around with this stuff and taking the time to do the measurements on spring rate, length, and suspension frequency.

Also, MASSIVE thanks to Scott for getting the Matiz springs made in bulk for us in North America. I bought sets for both my Insights as well as my friends Insights, as I think every 1st Gen still on the road should have a pair of these in the back.

With the Corolla springs on the front and the Matiz springs on the rear, the car is finally suspended, and we can now start playing with F/R balance and damping rates. Should be interesting!

I should note that I added the 17mm Prius torsion bar to the rear twist beam as well (pics below), but I'm still playing around with the mounting setup. I used some washers to shim the ends of the beam, and I think I should have just clamped the beam directly over the bar. There is a little noise coming from back there so I think the bar is slipping a bit. Will probably remove those washers, reef down on the bolts (pinching the torsion beam directly onto the bar), and call it a day.

One thing I am toying with is removing the front bar and only running a rear bar (keeping it soft-ish with something like the 17mm Prius bar, tho I have the stiffer 20mm Corolla bar to play with too). The significantly stiffer front springs will have gone a long way to increasing roll stiffness up front, to the point where the front anti-roll bar may no longer be necessary. If I can achieve neutral balance with no front bar and a small-ish rear bar, and chassis roll stays in check, I think this setup would be great. Increased front wheel suspension independence is a good thing, especially with an open differential, or even a torsen type, which I am thinking about (may be necessary if future IMA battery modifications go as planned). One of the classic recipes for lightweight FWD cars is the old VW Rabbit GTI. Many of those are still being raced, and it's not uncommon to see them running no front anti-roll bar and a very stiff rear anti-roll bar (and one rear wheel in the air when cornering hard). Tho some guys go the opposite way, or just run insanely stiff springs and no bars at all. Probably not a way to go for a street car tho. Time will tell if this sort of setup is applicable to the Insight.

 

[JulianEdgar]

Good stuff!

Don't forget my car is a little heavier at the front (so explaining your fractionally higher ride height) and with that increased weight, I don't think I could get away without having a front anti-roll bar.

 

[JulianEdgar]

How is the handling now that you've been driving it a few days?

 

[samwichse]

Is that 17 mm Prius bar a drop in (obviously you have to drill the beam)? Or did you have to modify it somehow?

What years of Prius is it supposed to fit?

 

[Tycho]

Julian: The handling has been really nice. With double the front spring rate, one would expect bumps to feel harsh, but because the car stays up off the bump stops, it feels "firm yet smooth". Large bumps jostle your body a bit, but they no longer cause your teeth to fall out. All the wheels stay on the ground all the time. That is the biggest thing. There's no more shuddering or shimmying of the chassis when it encounters bumps while going around a corner. 165 series tires never felt wider (another thing you noted).

Chassis roll is flat flat flat. I'm not sure what the camber gain or caster angle is (how much camber gain when the wheels are turned), but if there is any at all, the chassis could probably benefit from a little more roll. Now that the car is 1" higher in the front, the static angle of the lower control arms should be more conducive to delivering some camber gain. I will look at these angles when I am at the shop later today (measuring the height of the inner and outer LCA pivot points). F/R balance still feels a little biased towards understeer, but it has been snowy so I have not exactly been able to push things. That said, today I am going to work on getting the rear anti-roll bar to stop making any noise, trim the rear bump stops conically to reduce their rate and height a bit, and possibly remove the front anti-roll bar. It would be great to know what the torsional stiffness of the stock bar is, but I don't have the means to measure that here. I will probably be somewhere where I can jig something up soon tho.

Samwichse: The Prius bar is from the first gen Echo looking Prius. It is something like 90cm c-c stock, and must be cut down and welded to 80cm c-c to fit in the Honda beam.

I will say that drilling the holes is a bit of a trick, as the Honda beam slopes upward a little bit all the way to the center of the car, so you don't want to drill perpendicular to the surface of the metal, you want to drill perpendicular to the ground. I didn't pay a great deal of attention to this on my car and ended up with lower holes that were right on 80cm c-c and upper holes that were about 81cm c-c (what happens when you do the drilling after several hours doing other stuff on the car and are ready to go to bed). I then got to spend some time with the drill "honing" the holes out a few mm so things would line up well enough to get the bolts through.

The Toyota bars have 16mm holes and 13mm bolts, so there is a bit of play in there, I'm sure to deal with manufacturing tolerances. You just have to tighten the thing down super snug in the Insight beam so it doesn't twist around on its own. It has to be very securely sandwiched by the beam on each side, noting that the Honda beam has a slightly different space than the Toyota beam (which is dead on for the right clearance with parallel faces on either side of the bar ends).

 

[Tycho]

Nother little update.

Went out and cinched the rear torsion bar bolts down and that solved the clunking. The bar had sort of "settled in". I left the washers in place tho, as it would pinch the beam a lot without them. The washers are 2.5-3mm thick.

I checked the front A arm pivot points, and the inside pivot is now maybe 15-20mm above the outside pivot with the car sitting level on the ground. That means you're going to see camber gain in the first 3/4 inch of compression as the car leans, so a little bit of chassis roll no longer puts the car on the outer tread blocks. That's awesome.

On the rear end, I tested the torsional stiffness by jacking up one side from directly below the rear spring (jack under the trailing arm). With the 17mm Prius bar in there, the side that is being jacked up just barely kisses the original bump stop when the other wheel starts to come off the ground. At full compression, the combined roll stiffness of the Matiz springs plus Prius anti-roll bar is now equal to the weight of the entire rear of the car. That is a huge improvement over the stock springs with no rear bar.

I am getting ready to drive the car south for the winter, and may have access to Laguna Seca (maybe after some of the IMA battery mods, I can sneak in to one of the EV rallys there), among other places. I'm taking the 20mm Corolla bar with me and will likely make up some other springs to play with. I'll be right next to a bunch of import auto salvage yards when I'm down south, so it should be cheap to play with stuff. This will provide ample tuning opportunities.

Something I'd like to play with at some point is dropping the car back down to stock ride height, but with proper rate springs and trimmed bump stops. If the springs aren't way too soft and there's an inch or so of free travel before the bumps, this should ride reasonably well without any aero sacrifices. Camber gain will be lower however.

 

[JoelR]

Cool stuff guys! I'd be super curious to find out what the actual increase in torsional stiffness the Prius bar makes. As I explained in my dead sexy DIY Home Depot bar writeup, I found that +60% stiffer than stock felt great but I never tried anything higher (just 30, 45 and 60% stiffer).

Joel

 

[Tycho]

I need to measure the lateral frequency of the car. I also have a 20mm Corolla bar I haven't had a chance to install yet. One of these weekends I will swap them around and take measurements of none/Prius/Corolla and post here.

Having driven the car quite a bit now, I feel that the Matiz springs are still a little on the soft side. I have a set of Gaz shocks on order, and the extra compression damping will help a little, but the higher rebound damping (much more rebound than compression increase) will probably result in a lower dynamic ride height than the setup now with stock rear shocks.

Something in the 150-160 lb/in range would probably be more well suited, and would reduce the need for a heavy rear bar as you're gonna get some roll stiffness increase with the higher rate springs.

Oddly enough, the front struts do not feel like they are being overdriven trying to control a 200 lb/in spring. I get no pogoing, no top out, etc on the front, just a bunch of those things on the rear. Also, the OE damper pieces on the car say "SHOWA" on them, which probably accounts for their price and longevity (probably the only Japanese damper manufacturer who makes a phenomenal product). Too bad they can't be re-shimmed.