To learn and share about dampers / shock absorbers

[DaveW]

To return to vinuneuro's post, a "logarithmic-decrement" plot is a way of estimating the damping ratio of a lightly damped simple mass-spring-damper structure. It is unlikely to work (in general) for a multi-mode structure, & it would be something of a waste of time for a race vehicle (if multiple decay peaks are visible, then more damping is required, or the vehicle requires "fixing").

A few philosophical thoughts, if I may:

To repeat what I have said elsewhere, setting up the suspension of any vehicle is a complex business, so it is unsurprising that the process has been simplified using "rules of thumb". The mathematical basis for such rules is generally unsound, but it is surprising, perhaps, how useful they can be, at least within a focused range of applications. It would be wrong, however, to assume that such rules can be used "in reverse" to make fundamental statements about the physics of the vehicle. An example: here are axle-based frequency response functions of load per unit hub acceleleration for a real vehicle. The crosses represent the outputs of simple models identified from the measurements, with model parameters shown in the legends. Parameter values can be used to estimate, with errors of only a few percent, values for unsprung masses, sprung mass, sprung mass c.g. position, effective spring rates & effective damping coefficients. Everything, in fact, to compile an equivalent symmetrical model of the vehicle apart from pitch inertia & tyre stiffnesses. So, have I discovered a globally useful method of "modelling" the dynamic response of a vehicle? The answer is no, as demonstrated by this abomination, which contains the same response functions, with similar model fits, but for a very different vehicle.

Road vehicle engineers have a habit of using corner weights, converted to corner mass, and spring stiffness to compute estimates of axle-based "natural frequency". However useful the estimates are to the engineer (& I wouldn't question that), it is incorrect to refer to them as natural frequencies. In fact, axle-based natural frequencies do not exist for a 2 axle vehicle, & neither do axle-based damping ratios, despite any conclusions that might be drawn from the first set of response functions referenced above.

[Belatti]

-Dave Williams in an article said that most drivers prefer digressive valving because it offers more feedback and a little more responsive in corner entry, but linear valving gives more grip. But for linear valving the driver must commit to the corner to counteract the loss of driveability. Thoughts on this?

I leave others to comment - but I can't think I would have uttered the last sentence....

[/quote]

I haven´t read the article, a link would be nice. Apart from that fact, that can make any of my comments be out of context, I tend to mistrust those kind of "rules" like "linear valving gives more grip".

Yes, digressive can be used to increase low speed damping without making the damper hard as a rock at high speeds and yes more LS damping gives more responsiveness but then I have seen cars with very nonlinear motion ratios that needed linear dampers to acomplish something similar to this. What gives more grip depends on the package -tires, geometry, dampers, driver inputs-.

[Belatti]

Road vehicle engineers have a habit of using corner weights, converted to corner mass, and spring stiffness to compute estimates of axle-based "natural frequency". However useful the estimates are to the engineer (& I wouldn't question that), it is incorrect to refer to them as natural frequencies. In fact, axle-based natural frequencies do not exist for a 2 axle vehicle, & neither do axle-based damping ratios, despite any conclusions that might be drawn from the first set of response functions referenced above.

Sadly I was taught that way. One of the very old school persons I know that valves dampers for amateur series calculates an "axe based natural frequency" to get the numbers he needs to do his job, then only uses the damper position D curve (the only thing he knows and his old dyno would give him in a piece of paper) and thats it. He ignores all that we have been discussing here but his experience gave him relations only he understands -and that are kept in his head- that works in the track. 100% experience based rules of thumb and 0% science.

I have a tendency to question authority and learn the whys. Thats why Im here reading instead of valving dampers with a test and error aproach.

[speedsense]

-Dave Williams in an article said that most drivers prefer digressive valving because it offers more feedback and a little more responsive in corner entry, but linear valving gives more grip. But for linear valving the driver must commit to the corner to counteract the loss of driveability. Thoughts on this?

Most drivers perceive a loss of "support" of the front end when switching from digressive valving to a "true" linear shock. The judgment of handling for a certain corner, starts at the initial turn of the steering wheel on entry. When the "nose" of the shocks on the turning wheels is removed and put into a linear control curve, the car's handling judgment is perceived as lacking support (some will say the front end doesn't feel "supported" or "propped up") This is a perception problem that the driver has.
IMHO-It would be correct, if the last statement were changed to "...to counteract the perceived loss of drive-ability." I would disagree that the car actually losses drive-ability. It hasn't in every case I've dealt with, once altering and preparing the driver's perception of feel, in advance and the precision of the car's turn in ability isn't effected, though it may feel that way on the first entry he tries.

[DaveW]

I tend to mistrust those kind of "rules" like "linear valving gives more grip".

... and you would be correct. The article was written by a journalist following a relatively informal discussion (apologies, I can't remember where he published it). To be fair, I was asked to comment before publication, but took the course of least resistance (i.e. minimum change) with something I wouldn't have written in the first place.

To explain the comment a little, perhaps (because it was at least based on one of my beliefs), assume that an optimal damper set-up exists (in a dynamic sense) for a given vehicle/tyre combination. If the vehicle & tyres are essentially linear, then linear damping will mean that damping will be (close to) optimum over a wide range of input amplitudes. If the dampers are digressive, however, damping will be optimum only over a restricted range of input amplitudes - the vehicle will tend to be under-damped (for different reasons) for inputs that are both lower & higher than the "optimum" amplitude. Hence digressive damper settings will tend to be circuit-dependent, whilst linear damper settings will not (or, at least, be less so).

Dampers are, of course, multi-functional, & as a result the "style" of damping is likely to be different for different applications. A rear drive vehicle, for example, will probably have relatively linear rear dampers for a "road" circuit (where traction is important), but they may be (highly) digressive for an "oval" circuit, where traction & chassis control is not as important as other properties. On the other hand, the front dampers of the same vehicle might be digressive for a "road" circuit (where steering time constant & "feel" are important), and linear for an "oval" circuit (where maintaining speed through a turn is important).

In summary, linear dampers, when they have been optimised to a vehicle, will yield best overall CPL control over a wide range of inputs. Non-linear dampers will lose (some) ultimate CPL control, but will offer other benefits that can (& often will) lead to improved lap times. Having said that, not all drivers are equal. Some (a few, in my experience) are happy to experiment. They tend to gravitate to more linear damping styles, & also tend to be seen towards the front of a grid (for whatever reason).

[747heavy]

Delphi presentation about passive dampers, and some explainations what different damping curves mean for road car development.

http://www.vehicledynamics-expousa.com/ ... ierzek.pdf

[DaveW]

Delphi presentation about passive dampers, and some explainations what different damping curves mean for road car development.

Interesting presentation, 747 (again). I could argue with some of the details, but it is generally, I think, a good discussion of the difficulties encountered, and solutions adopted, by road car development engineers.

A few random thoughts of my own (for what they are worth):

1. The presentation divided damper "effects" into velocity domains. Whilst this might be valid on average, it is not in particular, because velocities reached by dampers will depend on how well a vehicle is damped (overall), on the magnitude & "shape" of surface irregularities, and on the speed of the vehicle. This fact is recognized by manufacturers, at least implicitly, because the suspension of a vehicle will often have different specifications for different countries. Perhaps the biggest differences can be found between USA & European spec. vehicles, but most vehicles developed in the USA, for example, have alternative "California specific" suspension set-ups.

2. "Comfort" (as opposed to "handling") is the over-riding consideration for road cars. Some will disagree with that statement, but I maintain an "ad hoc" data base of over 300 road vehicles tested on a multi-post rig. My estimates of heave mode damping ratio for those vehicles cover the range from 12 to almost 70 percent of critical. It is difficult not to conclude that development engineers will achieve what platform control they can, but that is very much limited by vehicle properties & by a "comfort" requirement.

3. "Comfort" is not a subject raised often by competition drivers. As an aside (& I am certainly not a race driver), my road vehicle happens to be fitted with Magneride, which has selectable default & "sport" settings. The sport setting is evil on road surfaces that are essentially smooth (not that many of these exist in the UK), because it excites engine modes & (I think) it excites a mode that is me sitting on a sprung seat. By way of contrast, the sport setting makes the car both more consistent & more "comfortable" when driving quickly along uneven roads. That might partly be due to clever suspension algorithms, but I have concluded much the same in the past with passively damped vehicles.

4. It is the case, I think, that a "good car" on a poor road surface can feel much like a "poor" car on a good surface. If a road car becomes "out of shape", for any reason, then a driver would be expected to slow down until he feels in control, or comfortable, again (perhaps matching his damping characteristics to the current road surface). That isn't a solution a race driver would find acceptable....

5. The ratio of tyre/spring stiffness for a road vehicle will probably be between 10 & 20. In contrast, the same ratio for a competition vehicle (leaving aside off-road vehicles) will probably be in the range 4 to ... 0.3. It would be unreasonable, I think, to suppose that "rules" developed for one class of vehicle would be applicable to the other. (Again as an aside, simply bolting slick tyres onto an unmodified road car would almost certainly be a mistake).

Overall, I would suggest that the referenced presentation discusses "rules of thumb" used by road car development engineers to achieve satisfactory performance for a particular application (& environment). Please read an earlier post for my views about "rules of thumb".

[747heavy]

I did not post the presentation of Delphi, in a attempt to present the absolute thruth.
As I´m sure, that this is one of the subject where people will not reach consens (at least not very easy, and unlikly on all counts).

I just put it here as an addition to the thread, and to let people draw there own conclusions.
But I agree, in general, with your sentiments presented above.
But to be fair, I can´t speak much about road car development and the goals as far as damping goes, but I would suscribe to the "comfort" is "paramount" statement.

The main reason I posted the presentation was, that it shows that two dampers which are similar in there "force vs. velocity" representation (graph), can/will show different bahavior in a car.
This shows the limitations of the "force vs. velocity" graph.
But I think we are in agreement here, that even if it is considered an "industry standard" it is a rather poor representation of the damper characteristics.

Noteworthy, at least for me, in the context of the thread is, I think, the notion to look at "energy dissipated" (the area of the force vs. displacement" graph)and it´s effects on the damper/car behavior.
Some damper dynos (few) calculate this value, which I find useful, as it is a quick way to compare different settings, but being sure, that we don`t compare apples and oranges, by changing the total amount of damping with different settings.

Another reason for posting the presentation, was the mentioning of different piston vs. foot valve settings and there effects, as we have disscussed this before, while talking about hysteresis, as well as showing the influence of different damping in simulation "quarter car".
This ties in with another thread on this forum.

That parts of the presentation have no/little relation to dampers in race cars, is ,I think, very visable in the statement about rebound/bump damping ratio´s.
As we see racing dampers (more so lately), which are using higher bump damping values then rebound damping values quite offten. Even some road cars show this tendency.

Damper architecture and price (very relevant for road car dampers)playing a role here, as does the stiffness ratio between tire/suspension, as you mentioned.
Unsprung weight is another factor worth keeping in mind, when it comes to damping "styles", IMO, as some racing series have some rather uniqe requirements here. (V8Supercars,NASCAR come to mind, but there are probably others as well)

So I hope some forumers, found something interesting in it, but by no means was it an attempt to present some universal valid rules, or the absolute truth of damping.
It was just a attempt to add to, and futher the discussion of the subject here.
I feel, that I have achieved just that.

Hope you have a nice weekend Dave

[DaveW]

I agree completely with your post, 747. Thank you.

I don't have too many issues with the Delphi presentation, remembering that it is aimed at road cars. One part that I do not find convincing is their "SuperProgressive" valve (which is, I think, similar in concept to the Penske "Regressive" damper). The presentation admits that the "SP system" results in a "Decrease of suspension deflection... avoiding hitting travel limiters", & I mention it because some race teams appear to think the concept will offer an advantage. So, Delphi proposed a style of damping specifically designed to limit damper travel that operates independently of damper position.....? Another item for my box of "Black Magic", I think.

[WilO]

A couple of links....

http://www.vehicledynamics-expousa.com/ ... /white.pdf

(A presentation on suspension measurement).

These folks build some decent dampers too....

http://dynamicsuspensions.com/ http://www.multimatic.com/index.shtml

[747heavy]

I agree completely with your post, 747. Thank you.

I don't have too many issues with the Delphi presentation, remembering that it is aimed at road cars. One part that I do not find convincing is their "SuperProgressive" valve (which is, I think, similar in concept to the Penske "Regressive" damper). The presentation admits that the "SP system" results in a "Decrease of suspension deflection... avoiding hitting travel limiters", & I mention it because some race teams appear to think the concept will offer an advantage. So, Delphi proposed a style of damping specifically designed to limit damper travel that operates independently of damper position.....? Another item for my box of "Black Magic", I think.

I´m not sure Delphis "SuperProgressive" valve is along the same lines as Penskes "Regressive " valve. ( more on the Penske system here: http://www.penskeshocks.com/files/racecar.pdf and http://www.penskeshocks.com/files/WEB%2 ... 2-2010.pdf )

I thought that Delphis "SuperDegressive" valve was aiming in the same direction, which is limiting the accl. of the sprung mass by means of decreasing/reducing the damping factor/coeficient at higher velocities.
This requires sufficient suspension travel to allow the spring/bumpstop to temporary "absorb"/store the energy of the impact.
I think the aim is to limit the transmissibility of an impact ( e.g. curb strike) to the chassis (sprung mass).
I can see that there are benefits with such an layout for certain applications (in racing cars).
I think inertia valves in dampers aim in the same direction, and I have used them, with some success, at times in Touring car racing. (to improve curb riding capability)

The Delphi "SuperProgressive" valve looks to me a bit "odd".
It´s like, "let´s wait and hope our problem goes away" if not we need "a lot of effort" to stay in control.
I think that this is a typical "comfort driven " road car set up approach (but I could be wrong).

We let the unspung mass "gain momentum/velocity" with low damping and then if we reach critical speed and risk that we crash into the "bumpstop" we throw heaps of damping on it, to try to stop it. (F=m*a)

I think in this case a position dependent damping approach would work as good/better ( like the "hydraulic bumstop" settings of most rally dampers today or the "long needle bleed valve" of older FOX moto cross dampers.

I think the "SD" damping valve is very road car / comfort driven, but again I maybe wrong with that.

Both systems will lead to an increase of "hysteresis" effects IMHO, which poses some challenges in it self.

[DaveW]

Perhaps my choice of phrase was wrong when comparing the Delphi & Penske characteristics. "Similar in execution to.." might have been better. Both "blow off" (the Penske version being the more extreme) & both increase force again at higher velocities (the Delphi version being the more extreme, perhaps). Anyway, the Delphi "SD" scheme does not deliver as stated, I think. This, for example, describes a true stroke limiting algorithm. It appears to be derived fairly directly from something we incorporated into the Lotus active suspension firmware in the mid 1980's.....

I first encountered "blow-off" damping characteristics in the mid 1990's, when Dynamic Suspensions produced an additional control for Touring cars. This produced a force reduction when the internal pressure reached a set value. It appeared to work well on a dyno, but not when fitted to a vehicle, where it generated unstable force oscillations. I encountered it next when an additional control was added to Champ/IRL DSSV dampers which simply limited force to a dialed up value (similar to the Delphi "SuperDegressive" scheme). This certainly worked as intended (my simple-minded interpretation was that it "gave up" mechanical control in favour of "aero" control for discrete, extreme events). The adaption was later offered to ALMS/LMP teams, & then to ALMS/GT teams & was received with some enthusiasm in both cases. I believe that Ohlins now offer something similar for TTX dampers, & the Koni "4-way" adjustable dampers are, in my view, actually conventional "2-way" devices with adjustable blow-off grafted on. They can work well too.

"Gravel stage" rally dampers have special requirements. They must be matched accurately to soft springs in order to optimize "grip", but they are also required to reduce vertical velocity to a manageable value when the vehicle lands after being airborne for a time (before the undertray is ripped off). My simple calcs suggest the initial impact velocity can be around 5 m/sec, but my customers tell me that is an under-estimate. The Delphi "SP" scheme could certainly find an application for that circumstance, I think. (Incidentally, it is worth noting that this specific problem is dealt with best by a "velocity", rather than a "displacement", related strategy).

In my mind, blow-off devices/schemes implement "survival" strategies, in the sense that they accommodate occasional extreme events without being forced to compromise a suspension set-up everywhere around a lap (or through a stage). The consequence is, I think, that their operation will be very track-specific. Hence they should be adjustable directly, a feature that is not available in some implementations....

[ubrben]

Perhaps my choice of phrase was wrong when comparing the Delphi & Penske characteristics. "Similar in execution to.." might have been better. Both "blow off" (the Penske version being the more extreme) & both increase force again at higher velocities (the Delphi version being the more extreme, perhaps). Anyway, the Delphi "SD" scheme does not deliver as stated, I think. This, for example, describes a true stroke limiting algorithm. It appears to be derived fairly directly from something we incorporated into the Lotus active suspension firmware in the mid 1980's.....

So basically the Penske "regressive" shock is a blow-off that allows more flow than a conventional blow-off such that the resistance (and therefore force) drops rapidly? Does the force then increase at even higher velocities due to chocked flow - or does another mechanism come into play?

Ben

[RacingManiac]

The line of semi-active damper that we are designing now at work for off-road, heavy vehicle have built in blow-off as somekind of pressure control. Mainly to help the damper survive in high pressure event that might otherwise exceed the pressure rating of the seals. The general strategy when we are tuning the blow-off is that it should envelope the performance range of the semi-active mechanism such that normal function of the system can be retained, but it effectively clips off the force higher speed range to control the pressure. But the ultimate pressure will still build once you get out far enough out of the curve. And that is more of a restricted flow issue(the flow curve effectively "orificed-out"...

[DaveW]

So basically the Penske "regressive" shock is a blow-off that allows more flow than a conventional blow-off such that the resistance (and therefore force) drops rapidly? Does the force then increase at even higher velocities due to chocked flow - or does another mechanism come into play?

I would guess valve characteristics, Ben, but I haven't looked inside one. Also, I haven't seen one on a rig as yet, although one of my customers has bought some explicitly to explore the potential of blow-off for his application. Here is a description taken from Penske's website, & some (unannotated) force/velocity curves (bottom of page 48).

Personally, I can't help thinking that a negative force/velocity region is a bad idea (a bit like a tyre with a partially negative force-displacement characteristic) & is a recipe for instability (which it was for the DS attempt).