fe1rx Ohlins Installation

[_Ryan_]

Originally Posted by fe1rx

This is a build thread, not a prescription. I offer up my approach as well as my solutions as food for thought. The approach (methodical, thorough, tested, documented) is probably the most important take-away. Everything made has been drawn, dimensioned, toleranced. Of course the spacer is repeatable, but it is only part of my complete "solution".

I realize that most people just want to buy a kit and have it work perfectly out of the box so they don't have to touch it again. My approach is that I would like it to work well out of the box and then I would like to improve it. That is not an imposition with a seasonal car, but it would be with a daily driver. Also, I have a certain combination of daily refinement and track manners in mind. Everyone is different in terms of how much of the one they are are willing to give up in the quest for the other. There is no perfect solution.

Obviously I design and make stuff for a living, so I have a few tools and materials at my disposal. But I don't make car parts for a living, so I don't need to rush my design to market or to convince anyone it is the best solution to their problem. I have no intention of marketing any of it. Some of it is to poke holes in marketing hype - of which there seems to be a surplus.

Sure. But it looks like a useful part for a problem I have today.

If I wanted to buy a kit that we knew would work well, I wouldn't have bought Bilsteins, or 8" springs in the place of 7".

[E82MSport]

For those of you that have only read the last page or two of this build thread I would encourage you to actually read the entire log. The amount and the quality of the information within is something not normally found on a forum.

Thanks again fe1rx for taking time to post all this great info.

[135]

This is an excellent thread - one of the best technical threads I've read. You're meticulous in your approach, you have documented it in a very thorough manner and it's well articulated such that a lay person can (mostly) understand it. Also, the diagrams are extremely useful to assist in visualising what your explaining. I had to "appreciate" quite a few posts to give you credit where credit is due. Well done!

My questions are mostly regarding your calculations - I'd like to understand these in more detail so I can apply the same to my own case. This topic interests me quite a lot so I've spent a lot of time formulating my thoughts and addressing a number of points that you raised, the result being my post is v-e-r-y long!


When discussing the rear springs, you refer to the difference between kg/mm and N/mm and how it can impact the conversion to lbs/in. I'm not sure what units the Ohlins spring rates are measured in but the Swift Spring rates are, as per your example for the Z65-228-120, specified as both 12.0 kg/mm and 672 lbs/in according to the Swift Springs website. 12 kg/mm converts to 117.68 N/mm, which then converts to 671.97 lbs/in, but, in your example, you've used 120 N/mm and 684 lbs/in. Since you mentioned the differences between the units of measure, I thought you may have just rounded the N/mm measurement so, when I saw that you referenced 684 lbs/in in the associated spreadsheet calculation and graph, I was a little confused. Am I missing some minor or obvious detail?


Also, while it wouldn't be too difficult to replicate, would you be able to attach the spreadsheets that you use for your calculations?

Originally Posted by fe1rx

What I would really like is to have a general solution to the question of “where does the lower spring perch need to be located on the front strut?” The answer to this is useful for several reasons:

1) to assess the viability of different spring options
2) to assess the impact of those options on wheel/tire to spring perch clearance
3) to assess the impact of a change to the top mount (i.e. a camber plate vs. an OE mount)

...

Static Spring Load can be determined in the following manner:

1) Measure the spring length at static ride height (140 mm)
2) Subtract from the free length to get the spring compression at static ride height
200 mm – 140 mm = 60 mm
3) Multiply the static spring compression by the spring rate to get the spring load
60 mm x 60 N/mm = 3600 N = 810 lbs. We could stop here but let’s carry on to check the result for logical consistency.
4) In addition, the strut (gas spring) supports 40 lbs, so the total “spring force” is
810 lbs + 40 lbs = 850 lbs
5) Multiply this force by the Front Motion Ratio to get the resultant of the spring force at the tire
850 lbs x 0.960 = 816 lbs
6) Add the Front Unsprung Weight to get the expected corner weight based on this spring compression
816 lbs + 115 lbs = 931 lbs
7) This compares favorably (within 1.5%) to the measured front corner weights (with correction for missing brake calipers and pads) of 917 lbs.

The usefulness of knowing the static spring load is that we know now that any spring we install must provide this load to provide static equilibrium. From its free length and spring rate, we can easily calculate the amount of spring compression (and thus the static length) of any spring.

The Static Spring Load was calculated at 3600 N (or 810 lbs), i.e. the Static Spring Compression (60mm) multiplied by the Ohlins Spring Rate (60 N/mm). You then used this 3600 N value in the spreadsheet calculation for the Ohlin springs. You then went on to use the same 3600 N value in the spreadsheet calculation for the Swift 65mm 7" 60 N/mm springs.

Since the Static Spring Load was originally calculated based on the compressed spring height of the Ohlins spring (i.e. 60mm), I thought the Static Spring Load would have to be recalculated for the Swift springs because, for example, the Swift spring at 178mm free length may only need to be compressed by 15mm, resulting in a 163mm static spring length at ride height. This 15mm compression would result in a Static Spring Load of 900 N/mm (157.61 lbs/in), which is far from the 3600 N/mm Static Spring Load used.

Also, I'm not sure whether the top mount height should factor into a comparative measurement, i.e. the 26mm difference in top mount height (70mm OE vs 44mm GC plates) could be considered against the Swift springs compressed height of 163mm (in my example) giving a comparative measurement of 137mm (178-15-26mm) against the 140mm OE Top Mount based measurement (200-60mm), meaning a relatively like-for-like comparison.
Again, I'm confused.

Why is the same Static Spring Load used for both springs?
Maybe another question is, why does the Swift 7" spring still need to be compressed by 60mm?
Do you think it would be useful to determine the values based on an equal (or equivalent) lower spring perch height, and thus an equal ride height, and calculate adjustments from that point? Perhaps, provided the Top Mount Height is the same.
If measurements are based on using the GC plates, which are 44mm high, and the preferred 326mm ride height requires a 372mm strut length, then doesn't that mean the lower spring perch location must always be 185mm and the static spring length must always be 140mm (plus 3mm for spacers / thrust sheets)? If so, why does your calculation for the Swift spring result in the static spring length being 118+3mm?
Does your 3mm thrust sheet measurement cater for thrust sheets both above and below the spring?

Originally Posted by fe1rx

Not that it makes too much difference but were there multiple fractions of a millimetre that were rounded for the OE top mount in the diagram above resulting in a 1mm shortfall between the component measurements (70+143+158) and the strut length (372)?

Originally Posted by fe1rx

For deviations from this ride height, the static strut length changes by the Motion Ratio times the change in ride height. Therefore a 10 mm drop in ride height would result in a 9.6 mm change in overall strut length and would require a lowering of the lower spring perch by 9.6 mm to accomplish this.

More generally, the static strut length can be expressed as a function of ride height. Conveniently this is valid for any strut:

Static Strut Length = 372 + (RH – 326) x MR

I can imagine placing a 10mm spacer between the chassis' strut tower mounting point and the top of the camber plate would simply result in raising the vehicle - but by how much?

Would the position of this spacer be outside the area affected by the motion ratio? Or is a spacer in this position now effectively part of the Top Mount Height and, therefore, requires the motion ratio to be applied to it? If the former, this would result in the ride height increasing by 10mm. Otherwise, if the latter, this would instead result in the ride height increasing by 10.4mm (conversely, a 9.6mm spacer would have resulted in the ride height increasing by 10mm).

Depending on the previous answer, to compensate, the spring could be compressed by, 10mm or 9.6mm (motion ratio applied), respectively, to return to the original ride height.

I see the positioning of the spacer in this manner is a simpler approach as it also raises the ride height in a similar way to the positioning of your spacer, although, I realise that it doesn't result in the same benefit of increasing the distance between the lower spring perch and top mount to reduce the possibility of coil bind, as does your approach.


One observation is that, while on-vehicle, it may be difficult to use the top of the steering knuckle as the strut datum, therefore, could the lowest strut thread be consistently used as the strut datum?

Granted, it might not be suitable in your case since you have added a 10mm spacer below this point but I suspect for practically everyone else this would be a suitable datum point.

At least, that's what I've been using. Specifically, since my focus is on the position of the lower spring perch, all my measurements are from the lowest strut thread to the underside surface of the lower spring perch.

Originally Posted by fe1rx

Shorter springs raise another concern though. They inherently have less usable stroke than longer springs. Too short and you risk bottoming out the spring before you achieve what should be full usable suspension travel. There are various approaches to determining what “full usable suspension travel” should be. A simple one is that the strut should have 50 mm travel beyond static ride height. For the Z65-178-060 spring, this implies a maximum stroke of 110 mm. The published usable stroke for this spring is 106 mm (maximum is 118 mm) so we are reasonably close to satisfying this rule of thumb. Incidentally, the front bump stop will begin engaging at 101 mm stroke, which appears acceptable to me.

I think it's interesting that you mentioned usable stroke of springs versus usable suspension travel. I think this is relevant to practically all suspensions makes - the one exception that I've come across is Moton (and potentially MCS, since they were born from Moton). The Moton coilovers were developed specifically for each vehicle such that (i) the maximum useable suspension travel (i.e. stroke of the strut shaft) is greater than the usable stroke of any spring combination (that could be fitted within the range of adjustment of the threaded body for the lower spring perch), meaning that the spring will always coil bind before the strut bottoms out, and (ii) they do not require bump stops, again, due to the aforementioned reason where the position at which the bump stop would normally be encountered/needed is not reached.
As a side note, some Moton coilovers (such as the ones I have) come with remote reservoirs, which are nitrogen charged at 175psi, and I've been advised that the reservoir pressure can be increased to preload the springs for better transitioning.

With my Moton coilovers and Z60-178-70 60mm 7" (178mm) 392 lbs/in Swift springs (Usable/Maximum Stroke 96/115mm) set with a static spring length of 163mm (Effective Usable/Maximum Stroke 81/100), I don't experience any coil bind. Nor did I experience coil bind when the static spring length was as short as 138mm (Effective Usable/Maximum Stroke 56/75) - this was purely based on no evidence of binding on the spring coils.

With all that being said, I would like to raise my lower spring perch to provide further clearance from my front tyres and, as a result, raise the vehicle height. My approach may be different to the majority in that my aim is not for a target ride height but, instead, for the height to be determined by the lower spring perch location (to provide wheel and tyre clearance) and the subsequent corner balancing of the car.

Side note: Coincidentally, though, my current front ride height (without driver) is similar to yours at 328mm (L) and 326mm (R). My front ride height was previously 331mm (L) and 323mm (R) when corner weighted (with driver) at 870 lbs (L) and 930 lbs (R) --- without driver, it was 326mm (L) and 318 (R), essentially 5mm lower (strange?) --- but I have since raised the ride height on the right strut to arrive at my current ride heights. As mentioned above, all my spring measurements use the lowest strut thread as the strut datum.
A further note to the above front corner weights, my car is a RHD so it can't achieve the more uniformly weighted F/R & L/R figures of a LHD car, although my diagonal was fine-tuned to 50.1/49.9.

To that end, rather than using a 7" spring that is preloaded, I'm trying to determine what the ideal lower spring perch location is so I can consider changing to solely a non-preloaded (or on the brink of being preloaded) 6" spring OR to a 5" spring combined with a Swift Helper spring - the latter option would ensure zero pre-load with the ability of the Helper spring to decompress at full droop to ensure the main spring is not dislocated or does not float.

Since the Helper spring doesn't add to the total spring rate, only the main spring rate needs to be considered. In addition, based on feedback from another 135i with Moton coilovers, I'm considering changing to a higher spring rate for the front springs. Therefore, based on the absence of, or potential for, coil bind, the two options I'm considering are:
- Z60-152-100 6" 560 lbs/in, Usable/Maximum Stroke 74/96mm
- Z60-127-100 5" 560 lbs/in, Usable/Maximum Stroke 68/78mm + Helper spring
For the Helper Spring, there are three options:
- H60-70-08 Free length 70mm, Fully compressed height 27mm, Stroke 43mm
- H60-60-15 Free length 60mm, Fully compressed height 19mm, Stroke 41mm
- H60-60-30 Free length 60mm, Fully compressed height 22mm, Stroke 38mm

While the lower spring perch location is yet to be determined, I suspect that it may raise ~10-15mm, although I'll use 10mm for any pre-calculations that I do. Therefore, based on my current 7" spring, the static spring length would be reduced to 153mm, resulting in the Effective Usable/Maximum Stroke being 71/90mm.

With these figures, it would make sense to use the 6" spring and raise the lower spring perch by an additional 1mm so the static spring length matches the 152mm 6" spring length, thereby, being at on the brink of preload. The spring could even be preloaded by up to 4mm more to achieve the 15mm increase in lower spring perch height.
This option provides little flexibility, though, as the lower spring perch can't be lowered in the future to reduce the ride height, otherwise, the spring would become loose at full droop.

When selecting the Helper Spring, it seems the one with the shortest FCH (i.e. 19mm) would be the logical choice, i.e. H60-60-15. With a main/helper spring combination, there is also the additional 3mm that needs to be considered for the extra thrust sheets plus an additional allowance for the main/helper spring adapter but I believe both of these would be absorbed into the further compressing of the Helper Spring by an equal amount.

If this Helper Spring was to be paired with the 6" main spring, then the ideal range for the static spring height would start at 171mm (152mm + 19mm FCH) and increase from there. We already know that the static spring length of the 7" spring is currently 163mm, which means the Helper Spring would need to be fully compressed and the main spring would need to be compressed by 8mm (and even more to cater for the additional allowance for the adapter and extra thrust sheets), so it would not make much sense to pair a 6" spring with a Helper Spring as a key purpose of the Helper Spring is to pick up the slack at ride height (generally partially compressed instead of fully compressed) and decompress to ensure proper location at full droop.

If this Helper Spring was to be paired with the 5" main spring, then the ideal range for the static spring height would start at 146mm (127mm + 19mm FCH) and extend to 187mm (127mm + 60mm Free length). At the current static spring length of 163mm for the 7" spring, the 5" main spring + Helper Spring combination would result in the 5" main spring being at its full free length and the Helper Spring partially compressed at 36mm out if it's 60mm free length (which doesn't seem as important) and 17mm (after having subtracted its 19mm FCH) out of its 41mm stroke (which seems more important). Again, the additional allowance for the adapter and extra thrust sheets would increase the Helper Springs amount of compression.

What I'm wondering initially, though, is whether the maximum usable stroke would be affected by the Helper spring. If you've considered this or have any formulae for determining this, it would be greatly appreciated if you could also attach them to this thread.

I know you've previously recommended against a higher spring rate but I thought those comments may have been based in the limitations of an Ohlins coilovers setup and, considering the absence of coil bind in a range of a scenarios with my Moton coilovers, I would be interested to hear your thoughts on the above options, especially regarding the 6" vs 5"+Helper decision?

[fe1rx]

Originally Posted by 135

This is an excellent thread - one of the best technical threads I've read. You're meticulous in your approach, you have documented it in a very thorough manner and it's well articulated such that a lay person can (mostly) understand it. Also, the diagrams are extremely useful to assist in visualising what your explaining. I had to "appreciate" quite a few posts to give you credit where credit is due. Well done!

My questions are mostly regarding your calculations - I'd like to understand these in more detail so I can apply the same to my own case. This topic interests me quite a lot so I've spent a lot of time formulating my thoughts and addressing a number of points that you raised, the result being my post is v-e-r-y long!

With all those "appreciates" I guess you have compelled me to reply! But I do appreciate that you are paying attention and keeping me on my toes. I am going to pick at this a bit at a time though, so that my reply is easier for me to keep track of.

Originally Posted by 135

When discussing the rear springs, you refer to the difference between kg/mm and N/mm and how it can impact the conversion to lbs/in. I'm not sure what units the Ohlins spring rates are measured in but the Swift Spring rates are, as per your example for the Z65-228-120, specified as both 12.0 kg/mm and 672 lbs/in according to the Swift Springs website. 12 kg/mm converts to 117.68 N/mm, which then converts to 671.97 lbs/in, but, in your example, you've used 120 N/mm and 684 lbs/in. Since you mentioned the differences between the units of measure, I thought you may have just rounded the N/mm measurement so, when I saw that you referenced 684 lbs/in in the associated spreadsheet calculation and graph, I was a little confused. Am I missing some minor or obvious detail?

I had that Swift data in front of me too. Strictly speaking they refer to "kgf" not "kg". A "kgf" (kilogram force) is not a normal metric unit, but in the same spirit as a "lb" (pound) is a unit of mass and a "lbf" (pound force, shorthand "pound" to confuse things) is a unit of force and is the force exerted by one lb when acted on by 1 g, a kgf is presumably the force exerted by one kg when acted on by 1 g.

In that scenario, the conversion from kgf/mm to lbf/in is kgf/mm x 2.2046 (lb/kg) x 25.4 (mm/in).

If you do the math 12 kgf/mm = 12 x 2.2046 x 25.4 = 672 lbf/in

That agrees with the Swift website data. The only problem is that it does not agree with my measured Swift spring rate data.

The correct metric unit of force is the Newton, which is the force acting on 1 kg when accelerated at 1 g.

1 Newton = 1 kg x 9.81 m/s^2 = 9.81 kg.m/s^2

Assuming "12 kgf/mm" really means "120 N/mm" we convert to lbf/in:

120 N/mm * 2.2046 lb/kg * 25.4 mm/in / (9.81 N/kg) = 685 lbf/in

That does agree closely with what I measured whenever I have measured Swift springs.

The "6.0 kg/mm" springs (stamped "60") should be 342 lbf/in if they are really 60 N/mm. I have measured between 344 lbf/in and 346 lbf/in. (3 springs).

The "12.0 kg/mm" springs (stamped "120") should be 685 lbf/in if they are really 120 N/mm. I have measured 684 lbf/in (1 spring).

By "measured", I mean I compressed the springs with a hydraulic cylinder and measured the resulting load with an electronic load cell. I measured the deflection with a caliper. I took readings every 10 mm (approx) and continued to a total compression of at least 100 mm. I plotted the results and used Excel to plot a linear regression line on the points from 20 mm to 100 mm. I rounded the slope of that line to 3 significant digits and that is the measured rate I am reporting.

It is too late to say "in short", but, in short the Swift catalog appears to be wrong (ie. produced by marketing not by engineering). I have to conclude from my testing that the Swift springs are manufactured in increments of 10 N/mm, not increments of 1 kgf/mm.

Incidentally, the Ohlins springs in the kit are 60 N/mm front = 342 lbf/in (I measured 344 lbf/in), and 70 N/mm rear = 400 lbf/in (I measured 399 lbf/in).

Originally Posted by 135

Also, while it wouldn't be too difficult to replicate, would you be able to attach the spreadsheets that you use for your calculations?

I have described the method so I will let you replicate it if you wish. I figure by the time you have done that you will understand it backwards and forwards. To be meaningful you will need some test data. Try this data for a Z65-178-060:

Force Compression
(lbf) (in)
0 0.000
147 0.374
250 0.681
355 1.004
451 1.295
552 1.606
652 1.906
752 2.197
851 2.476
950 2.756
1050 3.035
1152 3.327
1255 3.606
1349 3.862
1461 4.169

[fe1rx]

Originally Posted by 135

The Static Spring Load was calculated at 3600 N (or 810 lbs), i.e. the Static Spring Compression (60mm) multiplied by the Ohlins Spring Rate (60 N/mm). You then used this 3600 N value in the spreadsheet calculation for the Ohlin springs. You then went on to use the same 3600 N value in the spreadsheet calculation for the Swift 65mm 7" 60 N/mm springs.

Since the Static Spring Load was originally calculated based on the compressed spring height of the Ohlins spring (i.e. 60mm), I thought the Static Spring Load would have to be recalculated for the Swift springs because, for example, the Swift spring at 178mm free length may only need to be compressed by 15mm, resulting in a 163mm static spring length at ride height. This 15mm compression would result in a Static Spring Load of 900 N/mm (157.61 lbs/in), which is far from the 3600 N/mm Static Spring Load used.

Also, I'm not sure whether the top mount height should factor into a comparative measurement, i.e. the 26mm difference in top mount height (70mm OE vs 44mm GC plates) could be considered against the Swift springs compressed height of 163mm (in my example) giving a comparative measurement of 137mm (178-15-26mm) against the 140mm OE Top Mount based measurement (200-60mm), meaning a relatively like-for-like comparison.
Again, I'm confused.

Why is the same Static Spring Load used for both springs?
Maybe another question is, why does the Swift 7" spring still need to be compressed by 60mm?

To clarify, the "static compression" is the total spring compression from its free length when the car is sitting statically (not moving) on the ground. That spring is supporting its corresponding corner weight, less the unsprung weight, but it is doing so with a motion ratio which increases the weight supported by the spring relative to the sprung weight supported by that wheel. The static compression includes whatever preload you put into the spring when you installed it - perhaps that is what you are missing in your thinking? Every 60 N/mm spring will have a static compression of 60 mm at a load of 3600 N, regardless of its original free length. The only exception would be if you preloaded the spring to more than 3600 N, in which case it wouldn't move any further until you exceeded that preload.

Originally Posted by 135

Do you think it would be useful to determine the values based on an equal (or equivalent) lower spring perch height, and thus an equal ride height, and calculate adjustments from that point? Perhaps, provided the Top Mount Height is the same.
If measurements are based on using the GC plates, which are 44mm high, and the preferred 326mm ride height requires a 372mm strut length, then doesn't that mean the lower spring perch location must always be 185mm and the static spring length must always be 140mm (plus 3mm for spacers / thrust sheets)? If so, why does your calculation for the Swift spring result in the static spring length being 118+3mm?
Does your 3mm thrust sheet measurement cater for thrust sheets both above and below the spring?

A 178 mm (7") long spring (Z65-178-060) will compress 60 mm to a static length of 118 mm (+3 mm if you include thrust sheets). Swift thrust sheets are 1.35 mm thick each, so actually 2.7 mm if top and bottom.

Originally Posted by 135

Not that it makes too much difference but were there multiple fractions of a millimetre that were rounded for the OE top mount in the diagram above resulting in a 1mm shortfall between the component measurements (70+143+158) and the strut length (372)?

As you can see above 3 mm is actually 2.7. I figure accuracy to 1 mm is sufficient for these calculations as they really are approximate to get your spring perch location in the right ball park.

Originally Posted by 135

I can imagine placing a 10mm spacer between the chassis' strut tower mounting point and the top of the camber plate would simply result in raising the vehicle - but by how much?

This effectively increases the strut length by 10 mm. The strut has a motion ratio relative to the wheel because of its angle relative to vertical. I figure it is about 0.95. Therefore a 10 mm spacer on top of the strut top mount will raise the car about 10/.95 = 10.5 mm.

Something I omitted in my original approach was recognition that the OE top mount compresses under load. I have posted elsewhere that under static load it compresses about 5 mm. So if you are measuring things on the bench hoping to figure out your new ride height, you need to take that into account.

Originally Posted by 135

One observation is that, while on-vehicle, it may be difficult to use the top of the steering knuckle as the strut datum, therefore, could the lowest strut thread be consistently used as the strut datum?

Granted, it might not be suitable in your case since you have added a 10mm spacer below this point but I suspect for practically everyone else this would be a suitable datum point.

At least, that's what I've been using. Specifically, since my focus is on the position of the lower spring perch, all my measurements are from the lowest strut thread to the underside surface of the lower spring perch.

The Ohlins instructions locate the top of the lower spring perch from the very bottom of the strut - not particularly useful once installed.

They also reference the bottom of the check nut from the last thread - which is what you are suggesting.

In my case I wanted an exact (within 0.1 mm) repeatable datum point that I could use to preset the spring perches before assembling the struts in the steering knuckles. So I actually measure from the small lug on the side of the strut, not from the top of the steering knuckle (although they are theoretically in the same place).

After the struts are assembled, I am no longer interested in that measurement. Any change I would make from there would be based on ride height and corner weight measurements. That original measurement is just a starting point.

As a point of interest though, my initial front spring perch locations were accurate enough that I did not need to make any adjustment at the front. I did make some rear adjustments to achieve corner balance though.

[fe1rx]

Originally Posted by 135

I think it's interesting that you mentioned usable stroke of springs versus usable suspension travel. I think this is relevant to practically all suspensions makes - the one exception that I've come across is Moton (and potentially MCS, since they were born from Moton). The Moton coilovers were developed specifically for each vehicle such that (i) the maximum useable suspension travel (i.e. stroke of the strut shaft) is greater than the usable stroke of any spring combination (that could be fitted within the range of adjustment of the threaded body for the lower spring perch), meaning that the spring will always coil bind before the strut bottoms out, and (ii) they do not require bump stops, again, due to the aforementioned reason where the position at which the bump stop would normally be encountered/needed is not reached.

"Every" suspension needs a bump stop, which is to say a progressive rate spring at the limit of compression so that something does not crash (springs bottom, strut bottoms out, a suspension component contacts the chassis, or a tire contacts the chassis). The alternative is to run very stiff springs that will not be ideal under normal conditions.

"Ideal" is actually not that difficult to figure out based on undamped natural ride frequency - choose the one you want, and a spring rate drops out of that choice. With enough suspension travel, you could get away without a bump stop, but our cars don't have enough to do that. The Ohlins strut uses an internal bump stop that you can't see, but it is most definitely there. My early measurements led me to think it was a series of Belleville spring washers, but later more accurate measurement show it to be a conventional urethane bump stop.

On any given strut, you will be able to find spring (perhaps in conjunction with a helper spring) that will result in potential coil bind.

But to your point on suspension travel - that is the point. Usable spring stroke is how you get it. You need enough spring stroke to get the amount of suspension travel you need for the conditions you are operating under. It is very dicey to rely just on spring rate to avoid crashing your suspension. Without adequate suspension travel, one you will hit a big bump and break something.

Originally Posted by 135

As a side note, some Moton coilovers (such as the ones I have) come with remote reservoirs, which are nitrogen charged at 175psi, and I've been advised that the reservoir pressure can be increased to preload the springs for better transitioning.

I don't think you have that description quite right. Gas pressure results in a lifting force that will raise your car. It is equal to, in your case, 175 psi times the cross sectional area of the piston rod. Raise the pressure and you will raise the car a little bit, which actually takes load off the springs.

To be clear, when I speak of preload, I mean how much the spring is compressed when the strut is fully extended. If you are using helper springs, the preload of your main spring is zero.

Increased pressure in the strut also increases the seal drag a bit. Generally friction in a suspension is considered a bad thing, but you can also consider it to be extremely low speed damping, thus it reduces body motions under slowly changing load conditions.

Originally Posted by 135

With my Moton coilovers and Z60-178-70 60mm 7" (178mm) 392 lbs/in Swift springs (Usable/Maximum Stroke 96/115mm) set with a static spring length of 163mm (Effective Usable/Maximum Stroke 81/100), I don't experience any coil bind. Nor did I experience coil bind when the static spring length was as short as 138mm (Effective Usable/Maximum Stroke 56/75) - this was purely based on no evidence of binding on the spring coils.

Just to clarify - I would call 163 the preload length, not the static length.

The thing about a lot of preload is that as you unload the tire, as soon as the spring extends to its preloaded length the wheel suddenly picks up off the ground. This looks dramatic at the front of the car, but it results in a sudden weight transfer to the opposite wheel, which will cause an immediate understeer reaction. Ideally, you want your inside front tire to get very light at steady state limit cornering, but not come off the ground.

Originally Posted by 135

With all that being said, I would like to raise my lower spring perch to provide further clearance from my front tyres and, as a result, raise the vehicle height. My approach may be different to the majority in that my aim is not for a target ride height but, instead, for the height to be determined by the lower spring perch location (to provide wheel and tyre clearance) and the subsequent corner balancing of the car.

You need to consider the big picture. If tire clearance is the issue, this is a reasonable approach. You will be reducing usable stroke, but then your experience says that is ok.

Originally Posted by 135

…
What I'm wondering initially, though, is whether the maximum usable stroke would be affected by the Helper spring. If you've considered this or have any formulae for determining this, it would be greatly appreciated if you could also attach them to this thread.

I follow your helper spring discussion but won't try to dissect it. Your calculations generally make sense.

Usable stroke is defined by the main spring and is reduced by preload. Because a helper spring lets you run zero preload on the main spring, a helper spring does not reduce usable stroke of that spring.

That said, if you are using a helper spring, that means you have selected a shorter free length main spring. By necessity, that main spring will have a shorter usable stroke than a longer spring (of the same design family) with the same rate.

Originally Posted by 135

I know you've previously recommended against a higher spring rate but I thought those comments may have been based in the limitations of an Ohlins coilovers setup and, considering the absence of coil bind in a range of a scenarios with my Moton coilovers, I would be interested to hear your thoughts on the above options, especially regarding the 6" vs 5"+Helper decision?

Based on the undamped natural frequency of the car, I would not suggest a front spring stiffer than 70 N/mm. At that, your undamped natural ride frequency at the front is about 2 Hz (when tires are considered in the calculation), which is in the high range of normal for a sedan based race car. Suspecting that your car, like mine, does dual duty, that is really the upper limit of reasonable. If the car doesn't feel stiff enough at that, I suggest that something else is actually the problem. If it is roll stiffness, that should be addressed with bars not with springs.

[fe1rx]

There are a few different approaches to making this calculation and generally there are some hidden assumptions that can trip you up. The simplest approach is based on a measurement or calculation of static wheel deflection (x). The ride natural frequency is equal to 188 / SQRT(x) in cpm or 3.133 / SQRT(x) in Hz.

This assumes that your spring is linear rate, your motion ratio does not change with displacement, you are not riding on your bump stop, and spring preload is zero. Since at very least the last item is not true, some calculating is required to use this method. We can calculate the wheel static deflection assuming there was sufficient droop travel that there was zero preload, but it needs a lot of data not readily at hand (except that I have measured it):

FRONT:

1) 60 N/mm (342 lb/in) springs
2) Motion Ratio MR = 0.960
3) Wheel Rate WR due to spring = 342 x MR^2 = 315 lb/in
4) Wheel Rate due to bushings = 7 lb/in
5) Total front wheel rate = 322 lb/in
4) Total Corner Weight = 902 lbs
5) Unsprung Corner Weight = 115 lbs
6) Sprung Corner Weight = 902 - 115 = 787 lbs
7) Static wheel deflection= 787 lbs / 322 lb/in = 2.44 in
8) Natural Frequency = 3.133 / SQRT(2.44) = 2.00 Hz

This overestimates the ride natural frequency of the chassis, because it omits the tire deflection, which is about 0.41 inches assuming a tire rate of 2200 lb/in. Increasing static deflection to 2.85 in reduces the front natural frequency to 1.85 Hz.

REAR:

1) 140 N/mm (799 lb/in) springs
2) Motion Ratio MR = 0.563
3) Wheel Rate WR due to spring = 799 x MR^2 = 257 lb/in
4) Wheel Rate due to bushings = 23 lb/in
5) Total rear wheel rate = 280 lb/in
4) Total Corner Weight = 823 lbs
5) Unsprung Corner Weight = 120 lbs
6) Sprung Corner Weight = 823 - 120 = 703 lbs
7) Static wheel deflection= 703 lbs / 280 lb/in = 2.51 in
8) Natural Frequency = 3.133 / SQRT(2.51) = 1.98 Hz

This overestimates the ride natural frequency of the chassis, because it omits the tire deflection, which is about 0.37 inches assuming a tire rate of 2200 lb/in. Increasing static deflection to 2.88 in reduces the rear natural frequency to 1.84 Hz.

Hence my conclusion that 60 N/mm front and 140 N/mm rear are reasonable spring rates for our cars, if the primary objective is track performance (and if the vehicle has reasonable suspension travel and does not ride on its bump stops, etc.)

[John_01]

Nice to see your calculation of suspension tuning frequencies. Probably you can get a bit less pitching in the ride quality when the frequencies are better matched.

I don't know if this is a silly question or if you are waiting for someone to pick it up. Why are the corner weights in your post above different from what you had in post #37, as 873, 851, 803, 779 ?

[fe1rx]

Originally Posted by John_01

Nice to see your calculation of suspension tuning frequencies. Probably you can get a bit less pitching in the ride quality when the frequencies are better matched.

I don't know if this is a silly question or if you are waiting for someone to pick it up. Why are the corner weights in your post above different from what you had in post #37, as 873, 851, 803, 779 ?

Well, no, it is you keeping me on my toes. I just plucked out the most convenient weighing report without really looking at where the numbers came from. I have redone the calculations (edited above) for the following weight assumptions, which are actually more reasonable:

3450 lbs with driver
50.8% Left Side Bias
47.4% Rear Bias
50% Cross Weight

These assumptions give a front corner weight average of 902 lbs, and a rear corner weight average of 823 lbs. Obviously the calculation is load dependent, but these numbers are reasonable for my car on the track.

While these new numbers theoretically improve the pitching response, this really is a non-issue. As you obviously know in theory the front natural frequency should be a bit lower than the rear (meaning static deflection should be less at the rear) so that pitching response to a bump is minimized at some typical speed. This has some applicability for softly sprung, lightly damped vehicles because the suspension damping permits the body to go through several cycles before the motion is damped out (and you can only actually tune this perfectly for one specific speed - highway speed normally). For a stiffly sprung more highly damped vehicle (i.e. sports car or track car) the damping is sufficient that no significant pitching oscillation ever develops.

If the vehicle were undamped, a pitching oscillation would be excited at the beat frequency of the front and rear suspension acting together (i.e. the front - the rear). If the front were 2.5 Hz and the rear were 2.0 Hz, the pitch would develop at 0.5 Hz. That is one compete cycle in 2 seconds. With any reasonable amount of track damping, this would be damped out in less 2 seconds so less than one cycle. Hence tuning the front vs. rear for pitching frequencies is not really an issue for sports cars. Other considerations are generally more important (e.g. roll stiffness distribution, available suspension travel without over reliance on bump stops).

To elaborate on the above, I have dug out some comments from Milliken and Milliken's "Race Car Vehicle Dynamics". Their comments:

"Good ride in the passenger car sense also dictates the relationship between front and rear natural frequencies which is usually incompatible with roll stiffness requirements." (p 580)

and, they suggest that, based on experience, ride frequencies for non-aero sedan race cars should be in the range of 1.6 to 2.0 Hz with the front higher (p 605). "Front higher" is in direct contradiction to what is required to reduce pitching oscillations in lightly damped street cars.

and "To obtain the flat ride in a lightly damped vehicle it is important to have the static deflection at the front greater than the static deflection at the rear. With heavily damped vehicles, where the main ride resonance is well suppressed, this requirement is less important (race cars, for example)." (p 796)

[John_01]

The corner weight numbers look better now. Thanks for updating it.

Originally Posted by fe1rx

To elaborate on the above, I have dug out some comments from Milliken and Milliken's "Race Car Vehicle Dynamics". Their comments:

"Good ride in the passenger car sense also dictates the relationship between front and rear natural frequencies which is usually incompatible with roll stiffness requirements." (p 580)

and, they suggest that, based on experience, ride frequencies for non-aero sedan race cars should be in the range of 1.6 to 2.0 Hz with the front higher (p 605). "Front higher" is in direct contradiction to what is required to reduce pitching oscillations in lightly damped street cars.

It seems like they may be neglecting the use of a front sway bar. If using soft front springs together with a moderate front sway bar, then its possible to have high roll stiffness and ideally tuned pitch characteristics at the same time. That assumes we care about road car ride quality and chose the damping ratio for that purpose. In that sense, you seem to be clear about your objective which is totally reasonable.

Thanks for the great thread so far. The detailed info and your explanation has been very useful. I've swapped springs and sway bars a couple of times already and I'm still learning.

[fe1rx]

While recognizing that the Ohlins BMS M100 Road and Track suspension kit are actually intended for the E90 BMW 3 Series, not the E82 135i, I had the following nits to pick with the kit for our application:

REAR UPPER ADJUSTABLE SPRING PERCH

The rear upper adjustable spring perch (as delivered with my kit – call it V1) does not positively locate the rear spring. When I disassembled my suspension it was clear that the rubber spacer separating the adjuster from the OE spring cup was migrating around.


Ohlins V1 Adjuster

To address this, I have machined aluminum adapters that eliminate the need for both the rubber washer and the OE spring cup. The aluminum adapter is joined to the original threaded sleeve with Loctite, after first cutting the flange off the threaded sleeve. Just like the OE spring cup, my adjuster has a light friction fit in the mating chassis feature, so no movement is possible under load.


Modified Ohlins V1 Adjuster

By eliminating the need for the OE spring cup and by cutting off some of the V1 steel adjuster, I save about 1 lb per corner.


Modified Adjuster Installed

Just as I completed this modification, I noticed that the current Ohlins installation instructions, dated 2013-09-20, show that Ohlins has a modified “V2” version of their adjuster that is functionally identical to what I produced. They call for a rubber washer between the adjuster and the chassis, but I am going to run without one. It seems I just missed out on the new version when I bought my kit.


Ohlins Second Generation Installation


REAR SPRING LOWER PERCH ISOLATION

No isolation provisions were provided with my V1 kit to properly provide a mating surface between the camber arm and the rear spring. The result would have been metal-on-metal contact at a point of substantial articulation – completely unacceptable.

To address this, I machined a standard Powerflex urethane spring isolator with the mating taper and radius to sit properly on the camber arm.


Custom Lower Spring Isolator


Isolator Installed

The modified isolator worked flawlessly. Interestingly the Ohlins V2 kit now includes a similar 2-part isolator, so obviously they also recognized this deficiency in the V1 kit. The V2 isolator consists of a rubber washer and a nylon bushing. This new feature also appears in their installation instructions dated 2013-09-20 shown above.

C-SPANNER HEIGHT ADJUSTING WRENCHES

The Ohlins adjustable spring perches and their check nuts consist of 4 different diameters (different perch and check nut front and rear) and are all adjusted using a single C-spanner (hook wrench). These tools are finicky and easily slip of their respective nuts, resulting in bruising of the aluminum nuts, particularly when trying to adjust the rear spring perches. Anyone who has used these tools will have cursed them.

To address this I have made custom wrenches (laser cut from mild steel and (will be) cadmium plated) that better fit the rear nuts. An offset on the larger wrench means that by flipping that wrench over, the handle is indexed 15 degrees from its previous position. With the Ohlins hook wrenches the next increment is 30 degrees.


Ohlins C-Spanners & Custom Wrenches


Custom Wrenches do not fall off

The new wrenches work extremely well and do not tend to slip from the rear nuts. I am really happy with these. I had a few extras cut, so if any Ohlins users would like a set, let me know.

FRONT WHEEL AND TIRE CLEARANCE

The 200 mm long Ohlins front spring requires a very low setting of the lower spring perch. At the recommended nominal ride height (remember Ohlins actually intends this kit for a 3-series), this produces interference between the spring perch and the tire (OE wheel with 225/35ZR18 tire).


Interference at Nominal Ride Height


By raising the ride height by 15 mm positive clearance between the spring perch and tire results. This fit issue is worse with wider wheel and tire combinations, making it virtually essential that shorter springs be installed.


Clearance at +15 mm Ride Height

The Ohlins spring is 200 mm long (7.9”). As I have written previously, I am using shorter 178 mm (7”) Swift springs, a 10 mm spacer between the upright and the strut, and custom taller camber plates to accomplish a more favourable tire clearance while maintaining my desired spring preload (to avoid coil bind) and ride height.


My 7” Swift Swift Spring Clearance

Without some effort, the range of ride height that can be safely used with these struts is limited. This generally is an issue with our cars though, because of the tight packaging up front, so other struts have similar issues. Spring clearance will also be influenced by your choice of top mount / camber plate, which will influence the required spring perch height for a given ride height.

[chris82]

Really enjoying your thread and build. The close up photos are a huge plus; it allows me to see the details that I otherwise wouldn't have noticed on the 1's suspension without getting under the car myself

[HP Autosport]

We typically use Swift thrust sheets in place of the rubber isolators so the rear springs can move freely with less spring rate spikes.

If noise is a concern, the thrust sheets acts like isolators as well.

[E82MSport]

I most definitely would like a set of those wrenches, as the supplied are quite inadequate.

PM me.

[fe1rx]

I have shortened up the wrenches a bit to provide a better swing. They are now the same length as the Ohlins C-wrenches. I am not quite sure why I made them longer to start with. Edges are now finish machined with a corner radius, so they are now a decent tool.

[_Ryan_]

Originally Posted by HP Autosport

We typically use Swift thrust sheets in place of the rubber isolators so the rear springs can move freely with less spring rate spikes.

If noise is a concern, the thrust sheets acts like isolators as well.

I did the same. (when Vorshlag failed to comprehend that their e36/e46 parts weren't listed as model specific )
No complaints. Using Bilsteins means that my rear spring adjuster was already of superior design and doesn't introduce noise.

[HP Autosport]

Originally Posted by E82MSport

I most definitely would like a set of those wrenches, as the supplied are quite inadequate.

PM me.

A little bit of dry lube on the threads and keeping them clean of debris will make your adjustments much easier. Thrust sheets or bearings between the springs and the seas will also help.

[E82MSport]

Originally Posted by HP Autosport

A little bit of dry lube on the threads and keeping them clean of debris will make your adjustments much easier. Thrust sheets or bearings between the springs and the seas will also help.

Because my car is daily driven, even throughout the winter it's hard to keep the threads clean. I always clean as much as possible before I make adjustments but I also use a product from Boeshield called T9 to protect the threads.

Great product -
http://boeshield.com/why-boeshield/

I also have thrust sheets that help adjustment.

The issue is the design of the wrenches (not terrible but not that good). For the money paid for these dampers you would expect something a little better. I guess I'm nitpicking and at the end of the day I didn't buy wrenches to get quality dampers.

[fe1rx]

Originally Posted by HP Autosport

We typically use Swift thrust sheets in place of the rubber isolators so the rear springs can move freely with less spring rate spikes.

If noise is a concern, the thrust sheets acts like isolators as well.

I consider the issues separate.

I use thrust sheets on the top to allow the spring to unwind more easily (i.e. minimizing rate spikes). For this purpose thrust sheets are only needed on one end of the spring.

The bottom end of the spring sees more angular misalignment than the top because of the pivoting action of the camber arm. At full droop the spring is not fully seated on the camber arm. I suspect any spring without an isolator that protects both the bottom and the ID of the spring will have some paint worn off in these areas by the grinding action between camber arm and spring. Perhaps 60 mm springs fare better, but I am using 65 mm springs and they are a loose fit on the camber arm.

[fe1rx]

Originally Posted by E82MSport

Because my car is daily driven, even throughout the winter it's hard to keep the threads clean. I always clean as much as possible before I make adjustments but I also use a product from Boeshield called T9 to protect the threads.

Great product -
http://boeshield.com/why-boeshield/

I also have thrust sheets that help adjustment.

The issue is the design of the wrenches (not terrible but not that good). For the money paid for these dampers you would expect something a little better. I guess I'm nitpicking and at the end of the day I didn't buy wrenches to get quality dampers.

T9 looks like a good product for that purpose.

Oddly, the front and rear spring perch and check nuts are all different. If you wanted a "perfect" set of wrenches for this kit you would need 4 different ones. I agree that complaining about the wrenches is nit picking - I mean how often do you actually adjust them? But every time I have, the check nuts acquire a new ding, which is annoying.

[KevinK2]

Originally Posted by fe1rx

Needing a distraction while waiting for my new front springs I decided to investigate the OE anti-roll bars.

My plan is to start out with the OE anti-roll bars because just the spring change will result in a significant increase in roll stiffness. Also, I have no data to support a decision to change the bars. So while waiting for my new front springs to arrive, I decided to collect some data on the OE bars. That means more measuring and more graphs.

Having an actual front bar stiffness measurement allows me to construct and validate a simple mathematical model of the bar. The usefulness of this is that, if it works, will be to allow me to estimate the stiffness of other bars that I may consider but not have the opportunity to measure. The mathematical model is based on these assumptions:

1) The arms of the bar are much stiffer than the torsion section, so all significant deflection in the bar occurs due to twisting of the torsion section (clearly not valid for the rear bar, but looks reasonable for the front bar).
2) The bar is formed from tubular material with minimal material removal at any section, so every section has approximately the same cross sectional area (possible because the bar is hollow).

As a first approximation, the cross sectional area was assumed to equal the cross section of the bar at the pin holes (assuming no hole was drilled), because the lug is formed by fully flattening the tube end.

The critical dimensions of the OE front bar are shown below:

Attachment 1010214

The bar can be simplified for analysis purpose as follows by eliminating the tapered sections.

Attachment 1010215

The stiffness of the bar follows from these assumptions and this geometry and is within 4% of the measured stiffness. By adjusting the assumed constant cross sectional area appropriately the model has been made to match the measured data:

Attachment 1010216

…

The reason the calculated vs measured stiffness is within 4%, is just pure luck, thanks to one mistake calulating the value of J .

I bow to your excellent SAE quality write up, with excellent images to convey your thoughts ... one of the best write-ups I have encountered. But, we need to talk about some swaybar tech. In calculating the twist, you used J = pi r^4/4, but the correct value is pi r^4/2. So you have stiffened the calculated stiffness by a factor of 2. But even so, you were still close to the rate of the bench test, which the Puhn equation is based on. We ran ito this on the Mazdaspeed6 forum, where the measured rate was about 1/2 the calculated rate (you would have the same issue with the proper J used) :
mazdaspeed forum, Phate, sway bar rate calc'n vs measured value Lower on this page is Phate's bench test set-up.

What we found was at the load we tested at (250 lbs on a 414 lb/in bar) we found the bar deflected as a rigid body in the oem ~stiff rubber bushings. Phate measured the bar motion next to the support bushings, and determined the effect at the bar ends. This was about 50% of the measured displacement during the test! This was noted on the top of the next page.

I came up with an imperical correction factor to be applied to the Puhn results to account for oem bushing deflection: Kpuhn/R^ 1.35, based on just one data point.
R= span between bar end holes / span between bushing centers. In our case, R^1.35=2, for 50% of the Puhn stiffness.

Note that urethane bushings are stiffer, but not tested. And, it could be at higher loads the stiffness increases toward the Puhn "no bushing" calc'n. On my old GT6, that I built for SCCA DP + trip dcoe Webers, I built the adjustable rear bar, and made pivot bushings for the front and rear out of split aluminum, grease with never-seeze ... no squeaks, no friction, no deflection.

Finally, you must think of the difference between the Bench Test / Puhn formula, and being in a corner. The Formula reflects a single wheel hit on the road, where the center of the bar is at about 1/2 the bar twist angle. In a corner, the bench test displacement is split between the lever going up on the outer side, and down on the inside by the same amount. This means the center of the bar has no rotation. Bottom line is you multiply the Puhn/bench rate by 2X to get the stable corner rate.

BTW, how I got here? Someone on Phate's Mazdaspeed thread that I was working on, offered a link to this site where actual bench testing was also happening. That's when I saw the "J" error, and noticed the oem bushings being used, like we did.

.

[fe1rx]

Originally Posted by KevinK2

I bow to your excellent SAE quality write up, with excellent images to convey your thoughts ... one of the best write-ups I have encountered. But, we need to talk about some swaybar tech. In calculating the twist, you used J = pi r^4/4, but the correct value is pi r^4/2. So you have stiffened the calculated stiffness by a factor of 2. But even so, you were still close to the rate of the bench test, which the Puhn equation is based on. We ran ito this on the Mazdaspeed6 forum, where the measured rate was about 1/2 the calculated rate (you would have the same issue with the proper J used) :
http://www.mazdaspeedforums.org/foru...ml#post2779496 Lower on this page is Phate's bench test set-up.

You are correct. I got the answer I was looking for (incorrectly, thinking that bushing deflection and arm bending were relatively insignificant effects) so didn't dig any deeper.

Testing is definitive. I did the math to be able to estimate the stiffness for a hypothetical different bar. That is certainly suspect now. A factor of 2 correction for arm bending and bushing deflection isn't appropriate for a wide range of bar diameters, because bushing deflection will become more significant for stiffer bars.

The Puhn formula is only for a specific bar shape, never found in street cars.

Originally Posted by KevinK2

Finally, you must think of the difference between the Bench Test / Puhn formula, and being in a corner. The Formula reflects a single wheel hit on the road, where the center of the bar is at about 1/2 the bar twist angle. In a corner, the bench test displacement is split between the lever going up on the outer side, and down on the inside by the same amount. This means the center of the bar has no rotation. Bottom line is you multiply the Puhn/bench rate by 2X to get the stable corner rate.

That assumes that there is no jacking at the corner in question and the inside wheel's "up" matches the outside wheel's "down". Better, I think, to convert the bar stiffness to be a function of roll angle rather than one-wheel bump travel.

I am using OptimumG's "magic number" spreadsheet which takes care of that math.