Hi DTG Matt , It's very cool to see that Train Sim World 4 has brought suspension physics to the game. This is a very good point in time for me to point out a few things. For context, I'm doing the IMechE railway challenge 2024 and am responsible for the preliminary design of the primary and secondary bogie suspension systems. From today's stream, I'll cover the remainder of the explanation you were trying to get at. Yes, a train bogie has two suspensions systems implemented. Primary suspension is between the wheel axle (normally at the journal box) to the bogie frame. This supports basically all of the vehicle weight using a coil/helical spring and a hydraulic damper. Both parts are bought at a specification which allows for an underdamped response to forced vibration, meaning that the body is allowed to bounce to protect the materials it is made of and offer ride comfort - see the figure below. As you can see in the figure, an underdamped vibration is highly sinusoidal. It's very elastic, and the damper stabilizes the system (brings the amplitude/output to 0m over time) to alter the system's transmissibility such that the majority of forced excitation frequencies avoid the natural osscillating frequency of the primary springs. This is brilliant, protects the wheelset, protects the car... happy days... or is it? Short answer is no, because the primary suspension system deals with a multitude of vibration frequencies rather than focusing on one. This is why some mass-spring-damper systems adopt two or more springs in series rather than only parallel. This is because you can combine the different response characteristics held with different springs... or choosing another type of spring entirely: Secondary suspension, between the bogie frame/bolster and the bodyshell, exists to isolate the car body from low frequency vibration. It's a shock absorber. The above diagram illustrates a single-covolute air bellow which can act as either an air spring or a shock absorber depending on the specification and pressure. If you look back at the graph earlier, note how responsive the critically damped scenario is. A more critically damped (Zeta >= 0.5) suspension system brings the amplitude to 0 in the shortest amount of time with no fluctuation... no oscillating motion at all. With the air bellow, this can be achieved with or without the need for a hydraulic damper. Vehicles, in general, aim for a natural oscillating frequency of 1Hz - an unlikely excitation frequency which allows the car to just be cushioned. When selected effectively, the air spring has little bounce (vertical magnification), isolating any bumps in the track felt by the wheelset. Not only does this prolongue the fatigue life of the material and structures, it improves ride comfort in another way. Trains today, such as the Desiro ML, make use of PID (proportional, integral, derivative) flow rate controllers for the air from the compressor. This allows the digitally enabled suspension system to alter the height of the car, and the stiffness of the spring in motion according to the height of the spring (calculated knowing the relationship between pressure and volume), speed of the spring as it rises and squashes (stroke) by differentiation as a function of position and time. The computer can therefore be used to detect the vibration of the car vertically or horizontally, and stiffen the air spring accordingly... meaning the bogie can continue to be battered while the passengers feel little to nothing. The equations used to determine the mechanical properties of an airspring differ to that of coil/helical spring because it is the realm of fluid dynamics, hence the PID flow rate control model. So what does this mean for Train Sim World? Well it's uncommon for locomotives to use air springs for secondary suspension because they're often so heavy that the car would just burst the diaphragm of the air bellow. Air bellow load capacity changes with pressure, changes with volume, changes with action surface area, changes with heat capacity... it just changes but this is defined in the datasheet of the air spring. In Train Sim World, where a coil spring is used for the primary suspension, I expect to see the underdamped oscillating frequency as illustrated in the graph. Amplitude does not effect frequency, the bogie frame should continue to bounce near to the springs' natural oscillating frequency at a shallow/damped amplitude as shown in the graph. For secondary suspension, we have to largely ignore the axles for this to make sense because this is relative motion - the motion of the bodyshell relative to the bogie frame. In Train Sim World, I would expect the car body to undergo more of a squishing action - squishing the air bellow and the diaphragm pressure forcing the car back up and overshooting its free (not forced by excitation) height and then resting again at the free height (Zeta >=0.5 vibration response shown in the graph). If you were to take this somewhat simplified and short introduction to vibration, a study of machine dynamics, and look into it further then maybe Simugraph will be capable of animating the kinematics of the car continuously in real time using a set of the same parameters which are fed into programs such as ADAMS Car or MATLAB or Ansys...
Wow very technical writeup! So if I understand this correctly, you're saying: locomotives have rather bouncy suspension, and passenger cars have much more damping so should not bounce as much as the locomotives?
Yes and no. You need to identify the suspension springs and dampers used to get an idea. Ignore whether it is a locomotive or coach, what it is does not matter in the slightest. It's just mass, springs, and dampers. As I've said above for the case above, the bogie frame vibrates in a very oscillatory way under its own weight and the car weight, where g is still 9.81 (I don't expect to see instantaneous speed). The bellow is squishy and squashes in a stiff balloon way. It's a shock absorber. Relative to the bogie frame, the car body undergoes an absorptive 'sitting down on a space hopper' motion relative to the bogie frame.
Very interesting insight. Thanks for sharing! I had this on my table some moons ago in R&D. But in the automotive industry. Dealing with pneumatic suspensions on trucks and busses. That's why this rings a bell. Reading with attention, I´ve just figured that the basic physics are the same (Yah well ...), but some aspects seem to be a completely different cup of tea. We were always struggling to keep the unsprung masses as light as possible, it was like chasing the holy gral. And we had the tires to be taken into account. These by themselves are air bellows with rather unpredictable and empirical behavior under the most varying conditions like deforming due to lateral acceleration, temperature changes, CG relocation during braking and accelerating or all these factors together. And then the tire setup must match the air bellow suspension and stabilizer setups. That on a 40t truck with wide payload spread or a 4-axle double decker bus. I'd guess - I'm not sure - but it seems these aspects to a good degree are neglectable on train bogies?! But you will have other issues to deal with, right? That metal wheel on metal rail should come up with some really nasty and destructive vibrations if they´re not damped accordingly. I try to imagine a flat runner doing its destruction work or resonances causing severe metal fatigue.
Precisely. There is also a damper to cover the lateral vibration and inertia while accelerating as you may have seen with the 40T trucks or otherwise because springs and bellows will also vibrate in every direction since they're inherently multi-DOF.
It can be quite overwhelming at times. We've got to use lots of standard phrases to describe specific behaviours we see in experiments. I've found great videos actually which can help: Secondary suspension relative to the bogie frame: Primary suspension: Both together: Notice how the wheel is subject to lots of high impact, high amplitude shock and the primary suspension translates this into high frequency underdamped vibration. And then note how the car body and camera continue to move smoothly as the near critical secondary suspension takes away all of the low frequency shock. Both suspension systems together, with their different characteristics, tackle the response.
I can't wait to see suspension added to first-generation UK EMUs. If you've ever been on a Class 304, 310 or 312 crossing points at over 60mph in real-life, you'll remember literally getting bounced through the coach roof or into the nearest aisle. Those spring-loaded seats were like the 1970's sofa at your grandparents house.
That's the other thing. Driver seats and some older passenger seats have springs. Suspension is everywhere, even your backside!
ItsYa165 are you studying for a MechEng Degree, and planning to go into the railway industry or is there a Railway Engineering degree? I'm just curious, I'm hoping to the the Formula Student for 2025
I'm already doing both. I'm in train manufacturing and I'll finish my MechEng degree next year hoping to go into electronics masters.
If you're thinking about formula student, you can do it. I'd recommend IMechE railway challenge over it though if you want to do rail because the IMechE markets the challenge as the loco version of formula student. But I do the challenge on behalf of the company, not on behalf of the university.
Regarding the newest developments in suspension, they look very good. Remember that, at the track joint, the track itself is also subject to stroke like a leaf spring. In our physical world, everything with mass vibrates and moves. Variations in the stroke will have an impact on the stroke of the suspension systems onboard the vehicle - this is something which will be good to include if not already represented. Also note that any relative motion, caused by vibration, driven into the transmission system will impede the output or alter its load (moment of inertia). Lastly, older TSW stock would benefit from the new vibration physics.
I think you opened a can which DTG probably will realize its too much effort to implement all of it. In my opinion there has to be a mid way. A pacer without spring suspension feels off. But modeling all the suspension related thing in full detail? I struggle to see the path DTG is going longterm, especially after so many cutted content, minimal timetables because of "we just didnt have time". Nice to have a suspension physic, i just hope other features wont fall short because of that. Here to say the pacer feels real to drive, just a total different experience. Very good work!
You'll be waiting a while, as far as I'm aware it's only TSW4 stock and after that's getting the new Suspension system.
There are no first gen EMU's in game though. So it's just down to whether they actually produce one for the sim.
This isn't in my hands. But if DTG wants to improve their suspension physics, the only correct way to do it is to build a mathematically based suspension model which relates basic parameters (so it's easy to implement) - as expected for a simulator. The maths also isn't terribly difficult to be honest if you are an engineer or of the STEM discipline; Googling the equations of motion for translational, rotational, and torsional vibration of various spring types is less than a 1 minute task and you can take the equations of motion to calculate all sorts using good old-fashioned calculus. https://www.iitiansgateclasses.com/Document/13. mechanical vibration.pdf Formulae are universal, and from this sheet you can get position, velocity, and acceleration for any spring as long as you know or can guess its basic properties. But it's connecting it to an animation which is the most work, and I can appreciate that this isn't really in the scope of TSW4 if that's what you mean. If DTG really cares about delivering a realistic experience, then they should develop a standard procedure for modelling the suspension which would streamline all workflows going forward (which should already partially be there anyway). The other thing is that DTG's money isn't my money. So I'm not going to empathise financially with them over how much time it takes to make a spring bounce in a video game. If they want it, they'll make it happen. If they don't, they are free to pass on it.
