Dinan Front Sway Bar fitment issue with 1M control arms

[Cbadca]

Looking for some feedback here -

I installed 1M TRW front control arms along with Dinan fixed camber plates when I upgraded to Koni Yellows last year. Still on stock 135i springs. Initially installed Eibach springs but they were too low for a DD. Went back to stock. Nice DD and very common suspension upgrade.

Last month I tried installing a Dinan front sway bar for the obvious reasons ahead of a day at Laguna Seca.

It didn't fit?! What? Well I bought from a reputable company so lets go make it right, I thought

I went back to Dinan and after many emails confirming and reconfirming that I have a 135i Coupe and not a convertible or X-drive they said they thought it might be the 1M control arms and that maybe I needed their 1M sway bar but wouldn't guarantee fit.

Upgrading control arms on these cars is super common. The Dinan website has no warnings against the use of 1M control arms. Nor did Dinan offer to make it right. In fact they have ignored my repeated requests to exchange or return my sway bar. Very disappointing.

Is it me or has anyone else had this fitment issue?

thanks

[Papa Popov]

What was the issue? Was it hitting the lower wishbone? That whole kit seems kinda whack given that you have to modify the ABS/brake sensor bracket.

Why don't you just return it and get another 28mm bar? I can't imagine that the tubular aspect of the bar and the extra adjustability justifies price.

[Cbadca]

The Dinan sway bar is too wide - I went back to stock.

My goals for the post:
1) to see if anyone else had this issue
2) find out if I missed something in the install
3) call out Dinan for a serious lack of support

[Suprgnat]

Originally Posted by Cbadca

The Dinan sway bar is too wide - I went back to stock.

My goals for the post:
1) to see if anyone else had this issue
2) find out if I missed something in the install
3) call out Dinan for a serious lack of support

Middle sway is a rear bar.

[Cbadca]

Really?! Its crazy that I took the pic from the Dinan website today and it is exactly what was shipped. Thanks for the direction

[Suprgnat]

Originally Posted by Cbadca

Really?! Its crazy that I took the pic from the Dinan website today and it is exactly what was shipped. Thanks for the direction

You may have missed it in the body of the text.

[Cbadca]

I pulled wrong Dinan sway bar pic earlier. let me try again with updated correct picture of Dinan sway bar.

The issue is remains the same - the ends of the Dinan bar are not as sharply curved and flare out as factory bars. When installed the Dinan sway bar on both sides were in heavy contact with the steering boot pictured. There was no way to connect the sway bar to the tie rods without this contact. thoughts?

[spidertri]

Did you make sure they shipped you the correct part number swaybar?

[ZeroR]

To me, it looks like you need to use longer end links (probably those pictured with the blue bar in the middle). Are you using the stock end links?
I don't have that bar so I could be wrong.

[Driven5]

Direct from the DINAN web site:

This 2 hole bar can be mounted in 3 different positions with supplied end links that are left and right hand threaded for easy length adjustment.

The problem would seem to be that either you don't have the included (and required) end links, or you need to adjust them shorter.

BTW the "2 hole bar can be mounted in 3 different positions" thing is technically incorrect. It can be mounted in 4 different positional combinations. There are 2 symmetric, and 2 (mildly) asymmetric with 1 biased right and the other biased left.

[tsk94]

Originally Posted by Driven5

Direct from the DINAN web site:



The problem would seem to be that either you don't have the included (and required) end links, or you need to adjust them shorter.

BTW the "2 hole bar can be mounted in 3 different positions" thing is technically incorrect. It can be mounted in 4 different positional combinations. There are 2 symmetric, and 2 (mildly) asymmetric with 1 biased right and the other biased left.

They mean 3 different positions which affect the roll stiffness the bar contributes. Both sides on the soft hole (1), both sides on the stiff hole (2) and one side on soft and one on stiff (3). Technically the 3rd option could be swapped side to create a '4th' option, but it would behave exactly the same.

There is NO bias side to side by the way.

[Driven5]

Originally Posted by tsk94

There is NO bias side to side by the way.

Says somebody who has either never done the math for themselves, or done it incorrectly.

First, lets talk about the expected steady state cornering scenario. A sway bar is a simple torsion spring. It transmits a torque from one end to the other. So both ends see the same torque. However, due to leverage, the side with the shorter arm will react a greater force into the suspension than the side with the longer arm.

So using simple round numbers, there is 100 ft-lb of torque being provided to each end of the sway bar. And to exaggerate the effect, let's say that one arm is 1 ft while the other is 3/4 ft. This means that the suspension on the side with the long arm experiences 100 lb of force, while the suspension on the side with the short arm experiences 133 lb of force. Now while the same total force (roll resistance) is experienced regardless of which direction the car is turning, consider this:

Turning one direction, the outside tire is being pushed down with 133 lb while the inside tire is being pulled up with 100 lb. Turning the other direction, the inside tire is being pulled up with 133 lb while the outside tire is being pushed down with 100 lb. If you're familiar with suspension design and tuning at all, you'll understand that such an asymmetry DOES produce a biased effect.

Now moving beyond that, let's look at the less expected straight line effects. Again, thanks to the leverage (motion ratio) difference, here is only no preload across the bar at one singular ride height. During 2-wheel bump or accel/decel, the moving the bar ends the same distance will rotate each end of the bar a different amount, imparting a twist on it. This means there will be weight transferred across the bar even when no cornering is occurring.

Now what do you think happens as things get less ideal and more real? As the car is slowing on the way into a corner, or accelerating out of a corner. The effects will add or subtract depending on whether it's acceleration or deceleration, and a left or right corner. More advanced car development, especially in circle track, has been using this effect to their advantage for decades.

[tsk94]

Originally Posted by Driven5

Says somebody who has either never done the math for themselves, or done it incorrectly.

First, lets talk about the expected steady state cornering scenario. A sway bar is a simple torsion spring. It transmits a torque from one end to the other. So both ends see the same torque. However, due to leverage, the side with the shorter arm will react a greater force into the suspension than the side with the longer arm.

So using simple round numbers, there is 100 ft-lb of torque being provided to each end of the sway bar. And to exaggerate the effect, let's say that one arm is 1 ft while the other is 3/4 ft. This means that the suspension on the side with the long arm experiences 100 lb of force, while the suspension on the side with the short arm experiences 133 lb of force. Now while the same total force (roll resistance) is experienced regardless of which direction the car is turning, consider this:

Turning one direction, the outside tire is being pushed down with 133 lb while the inside tire is being pulled up with 100 lb. Turning the other direction, the inside tire is being pulled up with 133 lb while the outside tire is being pushed down with 100 lb. If you're familiar with suspension design and tuning at all, you'll understand that such an asymmetry DOES produce a biased effect.

Now moving beyond that, let's look at the less expected straight line effects. Again, thanks to the leverage (motion ratio) difference, here is only no preload across the bar at one singular ride height. During 2-wheel bump or accel/decel, the moving the bar ends the same distance will rotate each end of the bar a different amount, imparting a twist on it. This means there will be weight transferred across the bar even when no cornering is occurring.

Now what do you think happens as things get less ideal and more real? As the car is slowing on the way into a corner, or accelerating out of a corner. The effects will add or subtract depending on whether it's acceleration or deceleration, and a left or right corner. More advanced car development, especially in circle track, has been using this effect to their advantage for decades.

Thanks for explanation.

I previously assumed such, but never investigated or 'did the math' myself to confirm. I've been since told otherwise by more experienced drivers than myself that the effect is not asymmetrical - apparently this was not true and I've parroted what I've been told.

I did do the math, just now, and you were correct. My apologies. And good to know for myself going forward!

[Driven5]

No need to apologize. I'm just glad to be discussing with somebody who is willing to fairly consider new evidence that challenges their existing understanding.

It's something that even many 'professionals' and 'experts' still believe, and propagate, because even if they do the analysis they more often than not stop as soon as they see the total roll resistance being the same in both directions... Not taking it far enough to understand that how it's applied at each corner is not. In addition to that group, my guess is that many who are aware are more than happy to let their competition believe and propagate this fallacy. The effects are typically going to be relatively minor, possibly imperceptible to all but the best drivers. So nothing that would cause anybody harm, but also possibly enough to make the difference between 1st and not-1st.

It's kind of like people targeting 50% cross weights, when what they're technically looking for is usually equal F/R distribution on both sides. The cross weights thing comes from circle track using it as a baseline for measuring 'wedge', but it doesn't actually mean equal left and right handling balance. Fortunately, 50% cross weight is typically close enough to equal F/R distribution on both sides that it will again possibly imperceptible to all but the best drivers. Once again though, also possibly enough to make the the difference between 1st and not-1st.

Note that with the above tuning knowledge in hand, that equal left and right handling balance isn't always the fastest way around any given track, even when turning both left and right, since every track has more turns one direction than the other and some are more critical than others.

[fe1rx]

Originally Posted by Driven5

Direct from the DINAN web site:



The problem would seem to be that either you don't have the included (and required) end links, or you need to adjust them shorter.

BTW the "2 hole bar can be mounted in 3 different positions" thing is technically incorrect. It can be mounted in 4 different positional combinations. There are 2 symmetric, and 2 (mildly) asymmetric with 1 biased right and the other biased left.

I agree with the "you need to adjust them shorter" assessment.

Driven5, thanks for pointing out the asymmetry. In retrospect it is obvious but, like many, I stopped thinking after realizing that the torsional reaction is the same at both ends of the bar, thinking myself smart ...

[Cbadca]

good discussion -

bottom line is two fold
1) Dinan sway bar is shaped differently and even using the adjustable link doesn't fit -
2) Dinan has refused to do anything about it - they used to be great, now they are not - imho

be forewarned - buy a 1M/M3 bar if you want to upgrade after installing 1M/M3 control arms - do not mix and match!

thanks for your replies

[Driven5]

Honestly, the bottom line is that you still appear to be jumping to conclusions on something that was stated to "maybe" be the problem, purely out of a lack of having identified other possible causes, despite no actual proof of causation. The ends of the sway bar don't *have* to be shaped the same if it doesn't sit exactly the same... Which it apparently does not.

Perhaps you could provide some photos of the installation, validating both insight into its correctness to the instructions as well as highlighting the problems in question? That would be extremely helpful for actual troubleshooting.

I'd also love to hear a technical explanation of what exactly the M3 arms (or bar end shape for that matter) are supposedly doing to cause an unrelated portion of the sway bar to make contact with an unrelated portion of the steering.

[fe1rx]

Originally Posted by Cbadca

I pulled wrong Dinan sway bar pic earlier. let me try again with updated correct picture of Dinan sway bar.?

Your corrected picture does not put all of the bars right side up, which may contribute to the confusion, and begs the question - did you install the Dinan bar upside down?

The following image shows the bars correctly oriented as installed (looking foward), and makes it clear that the Dinan bar would need shorter end links so that the bends it shares in common with the OEM bar are located similarly oriented (and thus clear of the steering boot) when installed.



It is surprisingly easy to try to install a bar upside down ...

[san40k]

@Cbadca were you able to solve your issue ?

I've run into fitting issue with the same dinan sway bar on 2008 335i with upgraded m3 control arms:
- the sway bar starts rubbing on steering tie rod way before reaching 1/8" of the control arm
- the included end links included are way too long for this "misfit" application

[san40k]

got into same issue with Dinan - they don't want to help with "custom" setup, didn't provide a suggestion or recommendation. I'd think this is a more common setup these days for non M E90 335i/135i.

I've ordered shorted end links, but no idea if sway bar and whole setup will function as designed.