Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations KootK on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

A Survey of Guidelines for Specifying Dampers 1

Status
Not open for further replies.

ErikPSS

Mechanical
May 16, 2020
18
Thank you for the references in past threads -- the paperback introduction by Staniforth was helpful to get oriented, and the Millikens and Gillespie textbooks taught me to find CG, then correctly estimate longitudinal and lateral load transfers, and finally solve for critical damping. I also have a better understanding of this forum and hope to ask general conceptual questions, since I can now check my math against the textbooks.

I would like to ask about guidelines for initially specifying damper force curves. Woods & Jawad (1999) claimed that “dampers are the least well understood suspension components. There does not exist any textbook formulas to produce damper specifications which will ensure appropriate damping characteristics.”

However, the Millikens (1995) gave a sample optimum damping ratio for road-holding of 0.45 for a compact passenger car (p811), with the “coefficient in rebound being about twice that in bump” (p800), while Gillespie (1992) previously noted “typical shock absorbers are dual-rate with approximately a three-to-one ratio between rebound and jounce damping” (p157).

Woods & Jawad concurred: “A 3:1 ratio has historically been considered optimal, and remains a very prevalent ratio” to “help keep the vehicle CG low during corner entry. This is especially beneficial to vehicles with marginal camber control.” They further suggested to “Specify compression damping as a percent of critical wheel damping, say 20-50%” and “Accept values of critical body damping in the range of 70-140%.” They actually used 30% of critical wheel damping for compression damping, and multiplied by 3 for rebound damping.

Matt Giaraffa & Samuel Brisson of OptimumG (2005) gave fairly complex guidelines for damping: an initial damping slope multiplied by 2/3 for low-speed compression and by 1/3 for high-speed compression, and by 3/2 for low-speed rebound and by ¾ for high-speed rebound, resulting in fairly low compression damping as a percent of critical wheel damping and a rebound:compression ratio of 2.3:1.

Jim Kasprzak of Kaz Technologies (2012) borrowed heavily (diagrams uncited) from the Millikens, “Non-aero racecars, or those without ground effects, typically use damping ratios between 0.5 and 0.7,” but while he acknowledged that “Most text books recommend a ratio of three to 1, rebound to compression,” he “prefers the compression biased damping on both the front and rear.” Kaz also defines the digression point at 2in/sec.

Dennis Grant of Far North Racing (Autocross to Win) recommended “Set the shocks at 65% critical damping between 0-3in/sec. Somewhere above that, digress the shock off to ~30% critical to handle the occasional high speed whack,” although his calculator works somewhat differently.

Finally, Danny Nowlan of ChassisSim (2013) gave simpler guidelines: low-speed bump damping ratio of 0.5 for a bumpy circuit, with low-speed rebound of 0.5 to 0.7, and high-speed damping of 0.3 to 0.4. Obviously, this results in a very low rebound:compression ratio, but the low- and high-speed compression ratios match Giaraffa and Brisson quite closely.

Following this brief survey of the literature, I am specifying the front dampers for a non-aero car with marginal camber control on a bumpy circuit. I am inclined to simply specify low-speed compression damping at 50% of critical, multiplying by 3 for low-speed rebound, then digressing both off at 2in/sec to 30% of critical. However, this leaves compression damping as percent of critical wheel a little low at 25%, and rebound damping as percent of critical body at 150% -- outside the 140% that W&J recommend.
Left_wb5ahc.png


I am curious if anyone can share feedback on any of these guidelines. Thank you in advance.
 
Replies continue below

Recommended for you

Your curves would be very difficult to achieve with twin tubes as the low velocity damping needs to be the same in rebound and compression.

"The Shock Absorber Handbook" by Dixon is probably required reading. For what its worth on road cars we define the curves up to 1.5 m/s, and that is by no means the maximum damper velocity seen in practice.




Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376
 
They will be monotube, digressive piston, but unfortunately non-adjustable.

The motion ratio averages 0.63. The forces are specified for the damper itself.
 
For a monotube the ratio between low speed damping in c and r is fixed by piston rod area and the piston area

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376
 
Thanks Greg.

Dixon confirms "Typically the extension force is three to four times the compression force," and "In practice, the asymmetry of damping may also vary. At the ride optimization this may be distributed 20/80 bump/rebound, whereas at the handling optimum this may have shifted to a more equal 40/60" (p126). But he later claims "The scientific research literature is devoid of a good explanation of the asymmetry of dampers" (p132).

As above, I thought this was almost universally accepted to “help keep the vehicle CG low during corner entry. This is especially beneficial to vehicles with marginal camber control.”

Or is Dixon referring to passenger car applications, where damper jacking is more problematic?
 
Damper jacking is more of a problem for offroaders than passenger cars. You'll have to read your copy of Dixon for context, mine is thousands of miles away.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376
 
In the motorcycle world, we call it "packing", and it's not just an issue off-road, but also on bumpy pavement (which is an issue on all tracks that I ride on in Canada). Too much rebound damping relative to compression damping lets the damper compress easily over a bump and then doesn't let it stretch out on the down side, leading to the bike taking up that motion. It's an issue because it leads to loss of available suspension travel for soaking up the next bump, and it leads to loss of cornering clearance, and if it's unequal front to rear, it dynamically changes the steering geometry as you go over a bumpy section, which adds up to a whole lot of not good. So although one would "like" to use as little compression damping as you can get away with to avoid having a bump impact upset the chassis, you end up having to compromise, by adding in compression damping so that you don't run out of travel.

The best dampers have separately adjustable low-speed damping for compression and rebound, although the high-speed damping is in the shim stack. The not-quite-as-top-level dampers have some crossover or interference between the compression and rebound adjustments. An Ohlins TTX shock has externally-accessible damping cartridges, separate for compression and rebound and serviceable without taking the shock out of the bike. My Ohlins shock ... isn't quite that advanced. As for compression and rebound being separate ... the Ohlins fork damping cartridges in my own race bike have these functions as separated as they can possibly be. All of the compression valving is in the left fork and all of the rebound damping is in the right fork. (The valve stack is bypassed by a check valve when the motion is in the other direction.)

I don't know the mathematical relationships. I reckon there's more rebound damping than compression, but I've evened them out relative to Ohlins' initial recommendations, to counter bottoming. I've made adjustments in the field until the bike does what I want it to do under the conditions that I'm riding it.
 
Interesting, thanks Greg & Brian.

I think this finally clicked for me this morning.
Dixon cites Sugasawa et al. (1985) as having "found theoretically a damping ratio of 0.17 to be the ride optimum (minimum spectral energy of body heave motion) and a ratio of 0.45 the optimum for road holding (minimum tyre force variation). In more detail, the value found for optimum ride in ordinary driving was 0.16, a value of 0.43 to minimise ‘bouncy feel’, a value of 0.44 for road holding on rough roads, and a value of 0.71 for roll and pitch minimisation with control inputs."

So I read Sugasawa (attached) and Table 1 gives Optimum Damping Ratios for Various Running Conditions, such that roll and pitch are minimized at 0.71, whereas road holdability improves at 0.44. But this was for an electronically-controlled active suspension which could select the optimal damping ratio. For context, Woods explains, "With active suspensions, decoupling of the modal characteristics became possible. Designers naturally were unwilling to give up the performance gains when active suspensions were banned after the 1993 season. Developments in damper specification are occurring rapidly in all levels of motorsport."

Was the rapid development & proliferation of digressive pistons aimed to recover the performance gains of active suspensions?

That is, an optimal strategy would seem to use the low-speed (0-2in/sec) damping ratio of 0.71 to minimize roll and pitch, then digress off to approximate the lower road holding-optimal ratio of 0.44. Of course, even with a high-speed slope <0.44, this would be a rough approximation, as the suspension would still invariably be overdamped at mid-low speed (2-4in/sec) and underdamped at mid-high speed (5-8in/sec).
 
 https://files.engineering.com/getfile.aspx?folder=8eaa695e-f8f6-4a0e-a6d3-b2d8378af09d&file=Electronically_Controlled_Shock.pdf
This is curious:
The foundational work seems to be from Fukushima, Hidaka and Iwata (1983) Optimum characteristics of automotive shock absorbers under various driving conditions and road surfaces:
Fukushima_ahhc3w.png


However, their digression point is closer to 0.15m/s, about 6in/sec. This is substantially higher than the current 2-3in/sec recommendations.
 
A full 3d damper characteristic makes you realise how poor the f vs v graph is at representing real damper behavior. Even worse is when you measure the force in the shock and the velocity as it drives over the road. A proper dynamic model of a shock absorber needs something like 40 terms.



Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376
 
Thanks Greg.

As above, I'm targeting a symmetric damping ratio of 0.71 to minimize roll and pitch at low speed, then approximating a lower ratio for road-holding.

However, at what velocity should I digress off, considering a motion ratio of 0.63?
I visualized it at Dennis Grant's 3in/sec. I balanced overdamping at 4in/sec with underdamping at 10in/sec:
left3_gyy9b9.png


Fukushima's 6in/sec. Overdamping is terminal:
left6_kwyn6w.png
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor