Ten Pounds of Shit in a Five Pound Bag - Twin Turbo LS3/DCT Daily Driver

[cholmes]

After looking at the solid modeling, I had another thought: if you have room, it's worth placing the mount as high as possible on the strut housing. The resulting longer links will have reduced angularity at extremes of suspension and steering rack travel.

[shainiac]

That's a good thought, but I'm using the collar as the bottom of the coilover sleeve. I'm not sure just clamping alone can hold the vertical load of the spring without being backed up by the strut ears.

[cholmes]

Yeah, I did the same thing with mine, I didn't trust just the clamping either. Plus, the brake line bracket is directly above the shaft collar on mine.

If you do want the raise the mount, a short piece of tubing or even a piece of coilover sleeve under the collar would make an easy spacer.

[shainiac]

I think doing a 2-piece collar is definitely the right move. I redesigned it with a captured bolt and it's much easier to get a longer bolt and better clearance now.

I also checked the motion ratio on the strut. WOW, It's 0.94! The stock TII sway bar would be 15% stiffer at that MR than my 32mm ST front bar at the stock end link pickup LOL.

At a 0.94 MR, the bar acts 2.3X stiffer than OEM pickup, and 3.7X stiffer than the inboard pickup I was considering. It looks like I can go WAY softer on bar now.

To match the ST bar rate, I only need a 1.125" x 0.083" bar, or a 1.25" x 0.049" bar.
To match that wheel rate with the OEM pickup, I'd need a 1.5" x 0.065" bar.
For the inboard location I'd considered on the tubular arms, I'd need a 1.625" x 0.083" bar!

The softer bar should be easier on all hardware as well.

[cholmes]

That MR is the shit, huh! I really like maximizing the MR on the bar for all kinds of reasons: smaller bar / high MR is more effective because there aren't huge forces deflecting bushings like a big bar, it starts working quicker on initial turn-in, it's easier on the hardware, lighter -- it's just a lot more efficient in every way.

I had the 1.125" (28.6 mm) Racing Beat front bar on my car at one time with the stock end links, and switched to a stock 23mm bar (not sure which FC model had this, I've collected several from junkyards) with the endlinks to the struts, and felt no difference in the roll stiffness. This is confirmed by your math.

I never bothered calculating the stock and modified MR, as I was doing this change to fix endlink binding issues. I could see just by eye balling it that the MR would be quite a bit better, and I knew from experience on other cars that it was the right direction to go. I knew I didn't want to keep the RB bar, and the 23mm one seemed like a good first guess; I also had 21 mm and 22 mm bars I could sub in if I felt I needed to. The car feels pretty balanced, at some point I'll track it to dial it in.

I also feel, although I can't prove it mathematically, that the car is less upset by sharp bumps hit by one wheel and not the other, when using a small bar / high MR vs a big bar / low MR, at the same overall roll stiffness. It sure feels that way on my car.

[shainiac]

That's good to hear that the softer bar confirmed the math.
My current bar is also 1.125", not 32mm. I'm not sure if that's a good starting place or not.

Stock bar arm length is 7.5". I could do 3 eyelets and do a few different bar rates. As close as I can space the bolt holes before they start overlapping is +/- 15% rate. I think I could make another bar for $<150 in materials, so not a big deal to make a new one.

I would like to lower spring rates in general for better ride and not really sure where to go for spring and bar rates.
Currently have 450/350 #/in springs.
GT350R uses 240#/in front spring rates for a much heavier car. 450/350 is pretty standard for off the shelf FC coilovers, but lowered FCs have a subterranean RC. Not sure if people are compensating for the bad RC with heavier springs or what. My car's also ~300# heavier than most FCs.

[cholmes]

I mentioned this calculator a few pages back:

http://farnorthracing.com/autocross_secrets16.html

Keep in mind, you don't have to fill in every block on this calculator to get spring rates. To use it, first click "clear values", then type in values for corner weights, unsprung weights (that is per front wheel and per rear wheel), the front and rear motion ratios, and the target design frequency. Click once on the blue area to lock in the values. Then just look at "Results" and read out the ideal spring rates on the top right. That's it, nothing else needed to get the spring rates.

Very helpful: hover your cursor over the heading in each data block, it'll highlight yellow and a detailed info window pops up.

Design frequency: the author of this calculator recommends 1.8 - 2.0 hz for hardcore street cars, 2.0 - 2.2 for dedicated track day and autocross cars. Make the rear figure 0.2 hz higher than the front.

So for a quick and dirty look at it, I put in some figures that approximate my car. I know it's 3,000 lbs with me and about 1/2 tank of gas, 51% front and 49% rear. For this example, I assumed equal weight side to side, so:

Front corners: 765 lb each side
Rear corners: 735 lb each side
Front Motion ratio: 0.94 (I copied your figure)
Rear Motion ratio: 1.0 (scientific wild assed guess)
Front unsprung weights: 90 lb
Rear unsprung weights: 80 lb ( both unsprung figures are, again, SWAG)
Front design frequency: 1.8
Rear design frequency: 2.0

Then I looked at the Ideal spring rates in the Results section: Front: 288, Rear: 314

Tried it again with 2.0 / 2.2 design frequencies: Front: 368, Rear: 394

Play around with it, see what you think.
I will say this: although most people run front springs quite a bit stiffer than rear, both my street and road race FCs have virtually equal rates front and rear, no rear bar, and a reasonably stiff front bar. Nice balance, perhaps not completely forgiving of sloppy driving. The prior stiffer front / softer rear / big front bar combo had serious understeer; pretty idiot proof, though.

[shainiac]

Thank you for the info! I'd bookmarked that calculator a while back but had forgotten about it.

I scaled my car a couple years ago with the 5.3L+EFR turbos, S195 knuckles/brakes, and air-air intercooler. I think net weight of the W2A is pretty close to zero, but 6.2 and new turbos might be slightly heavier?

2022 weights
LF: 800
RF: 795
LR: 736
RR: 721

Total: 3052
Cross: 50.2%
Left: 50.3%
Front: 52.2%

For yucks, I also have a photo of the car scaled with the 13BT, stock TII trans/diff, big single turbo, W2A IC, and single exit 3” exhaust.
I’m not 100% sure that the scales were under the correct wheels, though.

 2015 weights
LF: 634
RF: 729
LR: 763
RR: 654

Total: 2779
Cross: 53.5%
Left: 50.2%
Front: 49.1%

I added 60# to each of the LS corner weight (driver weight) and entered it into the calculator. I also upped the unsprung weight. For each corner, the tire/wheel is roughly 50#, 30# of brakes, 30# of strut body, knuckle, control arm.
The antiroll bar info was a bit confusing. There’s no input for bar ID, so I used my spreadsheet to get a solid bar to match the same rate as a 1.125”X0.083” bar. 0.91” OD.
The roll resistance from the bar is 2X that of the spring, which surprised me.

I think 350# springs front and back will be a good starting point.

Also, as if I didn't already have enough ADHD projects, I’m working on swapping late model Toyota GR86 seats into my car. I picked up a driver-side seat on eBay for cheap and chopped the OEM slider off of the seat mounts.
I was able to match the OE brackets up to 3 of the 4 holes on the Mazda OE sliders, but the seat is scooted pretty far inboard and hard to slide. The Toyota seats have 4 shoulder bolts that hold the stock brackets/sliders in from the side and act as pivots for the tilt feature.
My plan is to scan the Toyota seat mounting points + Mazda OE sliders and have some brackets laser cut and bent. As-is, the seats sit a bit higher than the stock TII seats, but without a sunroof, it’s not too bad. I’m sure passenger side will be worse, since that seat sits higher from the factory because of the cat converter hump in the floor. Hopefully with everything scanned, I can check seat bottom clearance and drop them a bit more.
They are super comfortable and have way more lateral support than my 37 year old TII seats. There’s also about 10 different flavors of color and material options from 13 years of FRS, BRZ, GT86, and GR86. The GR seats I bought are black cloth and don’t look too bad, but there are some nicer color/materials. I've always been a fan of OEM seats from a comfort, safety and material quality standpoint. I street drive the car far too much to deal with fixed back buckets. OEMs also use higher quality materials that are more resistant to wear and UV in my experience.
If everything fits well after I make brackets, I may consider selling them for people looking to do this swap.

[Exidous]

I get it's not the same car but those turbos are porky.

8.8
DCT
LS7
A/C
PS
Full interior
18x10.5 square 285/30-18
Full tank

Cross weight came out really nice. I was surprised how big a different made.

[shainiac]

Yeah, it's porky. I'm guessing a lot of the weight is in the turbo setup. I can't find photos, but I had weighed the W2A setup and A2A intercoolers and both added about 35#.
One turbo is approx 15#, cast manifold 20#, wastegate 3#, plus some extra tubing. 3" downpipes go from cylinders 3/4 to to the bell housing. Dual 3" into 4" at the driveshaft and 4" to the bumper with 2 resonators and a 2.5" muffler+valve. I think I weighed the exhaust without the downpipes at around 40#, maybe more. TBF, the 2022 weigh was with my old 3.5" exhaust which had a cutout and a huge muffler. It was probably 10# heavier. 

So, twin turbo vs NA with headers is at least 100-120+#.
I don't have AC, which probably costs 30-40#?
On-board air compressor setup adds probably 25# also. Subwoofer and amp is 25-30#. Dual pump surge tank is also an extra 10-15#.
Interior is completely stock with 15-20# of Dynamat added to the doors, roof, and spare tire area. All factory sound deadening is still there also.

All the factory wiring is still in the car, minus the engine harness and some of the front body harness. I like having all the factory HVAC working as-is through the Logicon and can't be bothered to pull the dash and spend days to depin 5-10# of unused wire.

The sunroof delete, fixed headlights, and tubular crash bar saves some weight Id think, but nowhere near what I've added.
I've definitely not focused on making parts as light as possible, usually erring on the side of being overbuilt lol. 

If it was a track car, I'd be more bothered, but I like my creature comforts and have more than enough power. Un-shitboxing these cars isn't easy and weight=quiet in a lot of situations.
The BRZ seats I'm looking to buy will probably add another 10-20#, but much more comfortable and heated lol.

[Exidous]

Oooo heated seats....oh la la.

[shainiac]

I switched gears a bit and have been thinking about how to make the Magneride tuning make sense.
Because the shock valving can be "reactive", the tuning is kind of ass backwards. Without knowing force, the shock valving is an output, not an input like on a shock dyno.

Running a fixed mag coil %duty cycle should provide a linear fixed valving, but I don't think it would be digressive.
It will be tuned in a 3D map. If I set one axis to shock rate of change  (in/s) and the other to shock displacement (in from ride height, -2" to +2"), I think it'd be a big feedback loop.
Changing the valving by rate of change will affect the rate of change. I can set it to one %DC for  positive RoC and a different for negative RoC to have a 2-way shock, though.
+RoC would be compression and  -RoC would be rebound.

I suppose I could make it act like a 3-way and drop %DC over a ROC threshold to act like a blow-off, but I dunno how well that'd work.
For example, one %DC for +0-3in/s, then lower %DC from +3-6 in/s for bumps.
There will definitely be lots of fiddling to get it working.

I've probably posted this before, but this is the only dyno I've found of a mag shock: