Chassis and suspension building techniques and approaches discussion thread

[nicjasno]

Me and @Gray Gear got into an interesting discussion about this topic and i thought i'd make a thread for it, so we can expand on it and not pollute the originating thread with it.

I'd love to see the heavy duty chasis @Gray Gear ws talking about to continue this.

Here's one of my approaches making an acurate and strong front part of the car (my lego bmw e30 attempt):

Liftarm frames and panels are a great way to make everything stiffer in a lighter package. I can not recommend them enough.

[Gray Gear]

I'll post pictures of my current suspension setup once I am at Home. 

[romashkaman]

Hello, @nicjasno! I have a question. Does a usage of standard Lego steering links cause a slack while rotating suspension struts? Because I noticed this problem. I remember, your Mustang and Challenger had steering links built out with bionicle parts. I tried that setup a few years ago - that was a really precise and reliable setup, you are a true master of lego-engineering. I have to say, you inspired me a lot in building models (not only cars) with goals to make them robust and playable. Thank you in advance!

Edited September 23, 2019 by romashkaman

[Didumos69]

Interesting topic.

In my believe, the two main aspects of building a good chassis are 1) coherence and 2) rigidity, in that order, because you first of all don't want your chassis to come apart too easily. Rigidity is sometimes not even a requirement, the UNIMOG chassis for instance has flex by design. Then again, when a model stores a complex drive train with gearbox etc. and you want it to run smoothly, rigidity becomes equally important. Personally, I tend to strive for both coherence and rigidity.

Coherence - Coherence (the quality of not coming apart easily) of a model can be obtained by using form-locking. Form-locking, as opposed to friction-locking, means that the main forces your chassis gets to deal with, work orthogonal to the main pins and axles that hold together your construction. For example: When you have several connected beams spanning the length of your chassis, it is better to have them oriented with their pinholes horizontally than with their pinholes vertically. Otherwise your vehicle is likely to fall apart under its own weight. However, if the first line of pins or axles holding together your structure can easily work their way out over time, your chassis will still come apart. If this is a likely event, you need to form-lock these pins and axles too, which means adding form-locking in a 2nd degree / dimension. Likewise you can add a 3rd form-locking dimension. To understand how, I need to explain a little about 'assembly closure'. People that know my builds, will probably recognize that my constructs (read larger assemblies) are usually 'closed' or 'locked' with 'closure' pins or axles. The 'closed' construct is completely locked-up in all 3 dimensions and will only come apart by breaking pieces or after removing the 'closure' pins or axles again. This is also why I often have 3L pins with bush and axles sticking out of my main structures.

Rigidity - To avoid bending in your chassis it is important to have some kind of bridge or console running through the middle of your chassis from the front all the way to the back. The higher, the better. I would say at least 5 studs high. To avoid torsional flex in such a bridge, you have to make sure the beams setting up your bridge remain squared. This is where triangles and squared elements such as frames and panels come into play. This allows for stiff but narrow bridge / console. So making use of frames and panels adds to overall rigidity, but even more so when used in all 3 dimensions and even more so when form-locked into your structure. To obtain a structure which incorporates frames in all dimensions, I usually work with base layers: lengthwise beams with horizontal frames in odd layers and widthwise beams with vertical frames in even layers.

Edited September 24, 2019 by Didumos69

[nicjasno]

I strictly separate the drivetrain from the chasiss. The drivetrain needs to be completly removable. This way you can easily fix/modify it and it does not bind with the chassis/body.

Also i build my bodies as unibodies. So a complete shell, where the exterior is part of the load bearing structure. This way i maximise torsional rigidity and minimise weight. Unnecessary weight can be a models worst enemy.

[Gray Gear]

Here it is:

 

 

[Gray Gear]

Of course there are still some reinforcements missing, especially some crossbars to prevent flex from side to side, but you can put a 1L Bottle water on top without anything flexing or the shocks giving in.

[nicjasno]

Looking at the angle of the steering link really hurts. :P

What is going to be between the suspension? And why does it need to be so stiff? I am very curious.

[Mechbuilds]

My kind of thread! 

I'm currently working on @nico71's cralwer:
 

But i lack the long shock towers that the moc uses.. Since i'm using buwizz, i'm removing the gearbox for the winch and placing the buwizz where the battery box is.. But i need to get some kind of shock system from the normal small shocks.. 

I wonder if anybody has done this crawler and done this exact mod? If not then it's up to me to make it happen.. Rear axle has been a pain to add the springs to.. Not impossible but hard to make the frame strong enough so the yellow springs don't just twist and pop out of the studs they're mounted.. Same issues with front axle currently.. 

Can't give you pictures before i get an usb C adapter so i can attach my phone to the PC.. 

[Gray Gear]

2 minutes ago, nicjasno said:

Looking at the angle of the steering link really hurts. :P

What is going to be between the suspension? And why does it need to be so stiff? I am very curious.

Next to the engine has to be enough space to fit a detailled Turbo setup, maybe even a Sequential Twinturbo if I can fit it. I dont want the suspension to engage when I rest my hand a little on the model using the HoG. The model will be probably about 2.5 kg, with a little more weight at the back, so there will be ~1kg on the front axle. I just want a constant rideheight when playing with it.

[nicjasno]

3 minutes ago, Gray Gear said:

Next to the engine has to be enough space to fit a detailled Turbo setup, maybe even a Sequential Twinturbo if I can fit it. I dont want the suspension to engage when I rest my hand a little on the model using the HoG. The model will be probably about 2.5 kg, with a little more weight at the back, so there will be ~1kg on the front axle. I just want a constant rideheight when playing with it.

What's the point of suspension then?
I usually have a ballpark figure spring setup while designing and then adjust the final spring rate once the model is complete.

The suspension is fun to play with when you can get some lean into it.

Thing is, once you have used a HoG setup on a suspension with proper geometry (caster, kingpin..) you will never want to go back. Suddenly one has feedback in the steering, which feels amazing :)

Edited September 23, 2019 by nicjasno

[Gray Gear]

If you take a look at my NSX, I also used a twin shock setup in front, and a quad spring setup in the back (2 Hard springs and 2 soft springs per wheel.) Its just right, puwerful to not give in when playing with it, but giving in when the model is crossing an obstacle. I want it to be like that on this model as well.

[Didumos69]

14 hours ago, Gray Gear said:

Of course there are still some reinforcements missing, especially some crossbars to prevent flex from side to side, but you can put a 1L Bottle water on top without anything flexing or the shocks giving in.

So it has vertical cohesion, which is good, but when your model is going to weigh 2.5kg, you need some cohesion in other dimensions too. One thing I notice is that you are using quite a few lengthwise beams with their holes placed vertically. At this stage this will behave fine, but when your model gets heavier, the chassis will want to bend under its own weight. With the vertically oriented beams in your chassis, not bending will rely on friction-locking rather than form-locking. The bend-under-own-weight forces apply vertically and the pins holding the structure together are oriented vertically too. These forces only need to overcome the friction implied when pulling a pin straight out of a beam. When you use beams with their holes oriented horizontally, the bend-under-own-weight forces will apply orthogonal to the pins holding your structure together and we all know it is very hard to pull a pin out by pulling it sideways. This is what we refer to as form-locking.

13 hours ago, nicjasno said:

Thing is, once you have used a HoG setup on a suspension with proper geometry (caster, kingpin..) you will never want to go back. Suddenly one has feedback in the steering, which feels amazing :)

This is very recognizable! Even when you only add Ackermann geometry the handling gets so much more fun, even in very light-weight models (I know the model below has very poor geometry apart from Ackermann).

13 hours ago, Mechbuilds said:

But i lack the long shock towers that the moc uses.

You might consider assembling them yourself from the cheaper 2909c03 (Technic Shock Absorber 9.5L with Soft Springs) and non-LEGO springs. On doctor-brick.de the following springs have been suggested:

Edited September 24, 2019 by Didumos69

[suffocation]

Love this thread!

Longitudinal rigidity is pretty easy to achieve; I seem to struggle, however, with torsional rigidity. Case in point, this carrier can be held at either end with a 2.5 kg crane on top, total weight 4.5 kg, and won't bend a fraction of a millimeter. If, however, I try to twist it, it has quite some give. Obviously this doesn't affect peformance as under normal operation the model is never going to be subject to that kind of twisting, but I'd still like to figure out how to make it rock solid in all three dimensions.

 

[Didumos69]

59 minutes ago, suffocation said:

Longitudinal rigidity is pretty easy to achieve; I seem to struggle, however, with torsional rigidity. Case in point, this carrier can be held at either end with a 2.5 kg crane on top, total weight 4.5 kg, and won't bend a fraction of a millimeter. If, however, I try to twist it, it has quite some give. Obviously this doesn't affect peformance as under normal operation the model is never going to be subject to that kind of twisting, but I'd still like to figure out how to make it rock solid in all three dimensions.

I can think of 2 things.

1.) Use transversal beams that lock together the frames you have running along the outside of your main structure. By locking these frames transversally, there will be less room for play / slack in the transversal pins tying your frames to your main structure. That way the frames also add to torsional rigidity.

2.) Suppose you have a center column with a lot of torsional twist, but with properly squared side walls (with frames) like in your carrier. When you look from above or from below while you twist it, you will see that the upper and bottom sides of the side walls will slide / rotate a little past each other, longitudinally. To fix this you can add frames or other connections that prevent the side walls of your column from sliding past each other, both in the top side and the bottom side of your column. In my rugged supercar this is exactly what I did (also after being disappointed about too much twist) and it made a lot of difference (see image below). I see you already have this kind of frames underneath the front and rear axles, but I don't know about the mid-section or the top-side of your center column.

Edited September 24, 2019 by Didumos69

[nicjasno]

Thing is that ladder chassis like this have a natural tendency to torsionally twist, even when made out of steel.

That's why automakers went from body on frame constructions to unibody construction. Because a unibody by design twists less than a body on frame design.

In trucks and heavy equipment, this is factored into the suspension and is not an issue. In cars you want more precision and safety of the survival cell.

Edited September 24, 2019 by nicjasno

[Cumulonimbus]

Very recognisable, increasing the torsion stiffness of a frame with efficient part use is always a challenge. This site helped me to understand where to add structure and what effect it has: https://dsportmag.com/the-tech/chassis-tuning-torsional-rigidity/.

Especially these images from the aforementioned site, I found very insightful (in the order of of increasing torsion stiffness):

[Didumos69]

20 hours ago, nicjasno said:

The suspension is fun to play with when you can get some lean into it.

I can confirm. Where one suspension arm compresses, another can relax / expand. When taking a bump, the expanding one is typically the one opposite (diagonally) the compressing one. This is also more realistic.

3 hours ago, nicjasno said:

That's why automakers went from body on frame constructions to unibody construction. Because a unibody by design twists less than a body on frame design.

You are right of course and I admire the fact that you incorporate this approach in your models. It's very realistic. I grew up with my father's Meccano chassis and got stuck in center column chassis.

3 hours ago, Cumulonimbus said:

Especially these images from the aforementioned site, I found very insightful (in the order of of increasing torsion stiffness):

Indeed very insightful. They also reflect what I tried to say in my response to @suffocation.

Edited September 24, 2019 by Didumos69

[suffocation]

Thanks for the tips, lads

I tried the same twistaroo with my off-road tow truck and noticed much less flex. Upon closer inspection, I realised I had indeed used more horizontally placed 5x7 and 7x11 frames, so I guess I was more or less (un)consciously all paranoid in trying to keep it as flexless as possible given also the considerable amount of axles running back to (almost) front. Some old WIP pics:

The one thing that's missing, as you all pointed out, is a decent amount of triangulation to get that bricklike, Nineties IBM ThinkPad feel. Interestingly, the rear axle attachments do indeed have this, shall we call it "4D" reinforcement, and show no flex whatsoever in any direction. 

 

[grego18f]

Hey @nicjasno, I take profit of what you said on the other thread (that any geometry is possible in Lego) 

Some time ago, I tried to add the McPherson Strut on the 42056 set. This is what I obtained: http://www.brickshelf.com/cgi-bin/gallery.cgi?i=6479598

(Other pics here)

It had quite good behavior when the wheels were straight. With the wheel turned, the springs bent when a load was applied on the front. As I have seen you have much more experience than me on complex mechanisms like that, do you think it is possible to create a reliable MacPherson setup with the deep rims?

 

 

[nicjasno]

4 hours ago, grego18f said:

Hey @nicjasno, I take profit of what you said on the other thread (that any geometry is possible in Lego) 

Some time ago, I tried to add the McPherson Strut on the 42056 set. This is what I obtained: http://www.brickshelf.com/cgi-bin/gallery.cgi?i=6479598

(Other pics here)

It had quite good behavior when the wheels were straight. With the wheel turned, the springs bent when a load was applied on the front. As I have seen you have much more experience than me on complex mechanisms like that, do you think it is possible to create a reliable MacPherson setup with the deep rims?

 

 

I can see right away what the problem is. The angle of the spring is too extreme and the top pivot is a complete disaster. Use the small turntables like in the 240z build in the other thread.

[grego18f]

10 minutes ago, nicjasno said:

I can see right away what the problem is. The angle of the spring is too extreme and the top pivot is a complete disaster. Use the small turntables like in the 240z build in the other thread.

I agree with you for the disaster . But placing a small turntable will not help because of the depth of the rims (the turntable is hitting the rubber). Same finding for the angle of the spring, sadly. There surely is a solution, consisting of putting the steering pivot less inside the wheel, but it would need more space to turn... if Lego had produced narrower tires (closer to reality)...

[nicjasno]

The spring needs to be therefore mounted further away from the center of the rim.

My future plans:

Edited September 26, 2019 by nicjasno

[grego18f]

It is a Porsche 991.2 gt3 rs (or the 992, it will depend on when I finish it), so basic McPherson strut suspension... But I would like to use the 42056 rims, which are too wide for the scale.

BTW, a big scale Landy Defender?

Edited September 26, 2019 by grego18f

[nicjasno]

10 minutes ago, grego18f said:

It is a Porsche 991.2 gt3 rs (or the 992, it will depend on when I finish it), so basic McPherson strut suspension... But I would like to use the 42056 rims, which are too wide for the scale.

BTW, a big scale Landy Defender?

You can make a macpherson with the porsche wheels. Just can't have the strut inside the rim.

I plan to keep the lego landrover exterior and redo the rest properly.