W123 hydraulic motor mount trial & result
 

I got the silly idea that a hydraulic mount, as used in a few high-end cars would eliminate the cabin noise from my diesel, especially that bothersome 3000 rpm highway whine at 65 mph (great gear selection M-B geniuses!). The only one that jumped out w/ in-line studs was the 1997-2012 Corvette (& 2002-2004 GTO), which was surprisingly cheap ($12 ea), so I figured why not try. I haven't read that anyone else ever tried similar.

The biggest mod needed was to rethread one stud from M10 to M8 coarse to thread up into the M-B mount. I ASSumed the ~1" extra height wouldn't be a problem, and shouldn't a gushy hydraulic mount sink a bit. Wrong, I could barely close the hood and had to wire in the radiator shroud high to not hit the fan. I also had to jack the engine up to the max to get the mounts in, even after cutting both studs as short as I dared. It was also very hard to get the stud started since the mount rubbed the inner fender at that height, especially with one side slightly higher and tilted. I had to crow-bar the engine over to get it started, and also to drop the lower stud into the hole. I was so geared up that I found a stainless bowl that fit perfect, after cutting one side to clear the K-frame. I could fix the height issue by machining off ~3/4" on a spare set of mounts, but read below first.

How did it work? Terrible. The whole car body shook so bad I could hardly stand to even idle, and didn't get much better at highway speed. Initially, I left off the engine shocks, thinking who needs those w/ hydraulic mounts. Installing them didn't help. I also wondered if I had somehow changed the engine, like when I knocked off the fuel filter and might have got air in the hose, but cracking each injector until it bled didn't help. The problem appeared to be that the mounts are actually stiffer than the factory ones. I tested them on my hydraulic press and they need some force to compress, almost like they are solid rubber.

I finally gave up and re-installed factory mounts, w/ polyurethane fill, as before. But, this time I installed 1/4" backer plates underneath to support the poly since otherwise it will eventually put too much force on the thin sheet-metal underneath and crack it, as it did in my 1984 (earlier post). The engine now idles and runs much smoother, but no better than before the hydraulic mount fiasco.

Probably for the best since hydraulic mounts are a continual problem in Corvettes, as well as Lexus (different design). Owners regularly find them leaking brown gook and collapsed. Many change to after-market solid polyurethane mounts for hundreds of dollars and live with some buzz.

more photos.
And I had to take off that p.s. return tube while installing, which lost 1/2 qt. But it did barely clear the tube. Lots of things to interfere when you mod, as most hot rodders appreciate.

Thanks for the write up...good info.

It's confirmation for me that these diesels need soft mounts. I was shocked at how much vibration was transmitted when I did my initial urethane fill with too stiff of a urethane. I went through a long troubleshooting process finally installing new mounts that confirmed it was the mounts...which I then filled with a soft urethane that solved both the vibration and collapsing problems.

I didn't have the problem with the cracking because my installation required 1/2" spacers or I'd have discovered what you discovered with that too.

Thanks again for your work and documentation.

You probably wouldn't need the dampers with hydro mounts. I'd also look for ones from a diesel engine since they will be tuned to the different type of vibration.

I would suggest one from a mk2 or mk3 vw diesel, but those aren't set up to carry the same weight as the MB drivetrain.

-J

Great creative experiment. It was worth a try otherwise we'd never learn anything.

I worked on industrial machines in my day gig and we were always trying to kill vibration. While soft isolating pads often helped, stiffening the mounts would also help in some circumstances. The reason we found after a lot of trial and error was that peak resonant frequencies could be moved around by making a system heavier - like rigidly locking it to another heavy object, or making a flexible thing stiffer. This is because the resonant frequency of a system is the square root of the stiffness divided by the mass.

If you can move the numbers away from the frequency that they're being driven at like your 3000rpm point you can often diminish the overall vibration. It's kind of like pushing a kid in a swing at the wrong rhythm or trying to get a car to bounce by pushing it up and down at the wrong rhythm. So sometimes I've found you can get a surprise by stiffening or making something heavier which is counter intuitive. Damping is only one of the knobs we can turn down.

That being said, I killed a lot of vibes in my SD by turning in the rack damper screw (yes, not only idle but smoother highway cruising) and doing a valve adjustment. A big one was tightening my tranny mount too. You've probably done all that. The W126 has a different insulation setup anyway.

Way to go on the attempt. A negative result is still knowledge.

Static stiffness aside, any mount designed for a engine that isn't a 3cyl or a 5cyl is going to be extremely harsh at idle. The large amplitude of the vibrations are greater than what the decoupler is designed for. The mount will immediately hit IOD. Even worse, at 750RPM those vibrations are going to be at 12.5hz. This is right around where most hydro-mounts are designed to have the highest dynamic stiffness in order to combat smooth road shake.

http://i.imgur.com/LjAcrJx.jpg

Bill, good post. But I think you need to find a hydro mount that will take the weight of the 617. Your talking around a 600 lbs om617 vs a 400 lbs LS3. It would be interesting to see something that would work.


.

I have been trying to improve machines for 50 years.. particularly when a part needs replacing... people make fun of me for trusting and appreciating the FSM instructions on these cars... for having some trust in the millions and millions of dollars of R and D which MB has invested.... and their general ' German ' OCD when it comes to metals and machines in general....
Sometimes their FSM instructions are not followed accidentally.... sometimes on these types of vibration items... they will specify two different firmness rates on the same item... like at the front and the back... the driveshaft has this on some models... and sometimes people will not know to order both of them since they look the same to the eye front and back... those Germans were really good at what they were trying to do... no stone left unturned.... many say that this many years later ' stuff ' has improved... sometimes yes.. sometimes not...

Don't the 60x engines have hydraulic mounts? That's what I first thought when I saw the thread title.

Yup, the same PN fit 4, 5, and 6 cyls.

IIRC the 126 applications have rubber mounts.

Sixto
83 300SD

Yes, my SDL has a similar setup to the 5 cylinders... rubber mounts and external dampers.

-J

Seemed the opposite problem. The Corvette mounts didn't deflect noticeably under the engine weight, and seemed much stiffer than M-B ones when I played with them on my shop press. I think the natural frequency comments are the best explanation, and validius post seems spot-on since it was a low frequency that seemed to shake the whole body. One can't hear 12.5 Hz, but you would hear the harmonics and sounds from parts impacting. No loss, I have other old cars I might try them on someday.

I once generated a similar vibration at idle. In securing my 1984 300D's detached-at-top K-frame, I ran a heavy chain axially back to the tranny cross-over. That chain vibrated at idle and shook the whole body at a similar low frequency that was very annoying. I stopped it by jamming some big sponges between it and the tranny.

do you still drive with sponges jammed?

thanks

As Mark noted, the V8 hydro mounts are tuned to a very different frequency than the vibrations from your I5.

VW sold I5 diesels, I know in the Audi 200 at least they used hydro mounts. Perhaps those would work better?

-J

No, that chain was a temp fix just to drive the 1984 home, but the big sponge did get rid of the annoying vibration. I re-welded the K-frame at home. I didn't relate any scenario with these hydraulic mounts and sponges.

Unlikely there is such a "tuning" for the mount itself. The critical dynamic properties of the mount are spring constant and damping. Whether it works for a V-8 or I-5 engine app depends on the whole system. The main advantage of a hydraulic mount is the much higher damping. But, unlike the simple spring-mass model in Physics I, hydraulics is very non-linear, increasing greatly with the amplitude of the motion. I didn't notice these mounts deflecting enough to do much damping. The engine itself didn't appear to move much at idle, but some attached components did, like the AC compressor suction tube was shaking wildly and I could damp it with my hand. I tried different idle speeds, but it was pretty bad from 500 to 1500 rpm, and even on the highway. As mentioned, I think the mounts were just too stiff and transmitted everything from the engine into the car's frame.

Not quite.

Hydromounts are designed to act as a mass-spring system internally. They are designed specifically in such a way as to avoid viscous damping.

Because it is full of wonderful pictures, i'm going to cite this study: https://etd.ohiolink.edu/!etd.send_file?accession=osu1240664127&disposition=inline

Take a look at Fig 2.4. Inside the mount you have:

• The top element. This provides the static stiffness of the mount.
• The inertia track through which the glycol flows as the mount flexes
• The rubber bellow which is an extremely compliant rubber membrane that contains a reservoir of glycol
• The decoupler. I'll explain this more in a bit. For now, pretend it has 0 compliance.

As the top element moves, it pumps fluid through the inertia track. The mass of the fluid in this inertia track can be modeled as a mass-spring system with the top element above it and the bellows below. Once again, it is designed to minimize fluid losses.

The behavior of a typical non-decoupled hydromount (or in this case, a decoupled hydromount that is past its IOD but more on that later) can be roughly seen in Fig 2.12b. Dynamic stiffness (k*) is represented by the hollow squares. Phase angle (the difference between the force input and the motion, also known as damping) is represented by the solid dots.

• 6-7hz: The input frequency is before resonance of the inertia track. The presence of the fluid makes the mount softer.
• 9-20hz: The input frequency is at or around the resonance of the inertia track. This leads to the fluid's inertia fighting against the input frequency. Dynamic stiffness peaks here. Phase is at it's highest. Fig 2.15 shows this nicely (lower graph)
• Above 20hz: Phase drops off. What is actually happening here is that the inertia track is 'locking up'. When the input frequency exceeds the resonance of a mass-spring system the mass simply stops moving. The stiffness now is simply the bulge stiffness of the top element. That is to say, it is acting like a thick rubber balloon full of fluid that has nowhere to go.

It is worth noting that because of the inescapable fluid losses that do occur dynamic stiffness of a nondecoupled hydromount INCREASES as the amplitude DECREASES. This makes for lots of harshness at idle which tends to be small amplitude, medium frequency vibrations. Go back to Fig 2.4b. To solve this, a decoupler is added.

The decoupler moved up and down within the limits of it's cage to allow fluid to avoid passing through the inertia track as long as the input amplitude is not so large that the decoupler hits the limits of its travel. Unfortunately, there isnt a good diagram on that study that shows this. The effect is that at small amplitudes (0.1mm) the dynamic stiffness of the mount is very nearly the same as the static stiffness. At higher amplitudes, (~0.5mm) the decoupler is well past its limits and the peak phase angle (for the given amplitude) is able to reach its maximum (across all amplitudes). Above this, the mount behaves like a nondecoupled hydromount. The amplitude at which the peak phase angle is halfway between 0 and it's peak across all amplitudes is commonly referred to as the Initiation Of Dampening (IOD).

Almost all hydromounts are designed for maximum damping between 10 and 15hz. This is a frequency where a phenomenon called 'smooth road shake' is often observed. On an apparently smooth road, the driveline can begin bouncing on the mounts at the system's resonant frequency.

OK, enough theory. Going back to my previous post in this thread, what i said was that compared to a V8 gasoline engine, the amplitude of the vibrations that an OM617 will create easily exceed the IOD of the mount (likely around 0.3mm, peaking at no more than 0.6mm). I also said that given the first order imbalance of an I5 and an idle of 750RPM, the frequency of the largest amplitude event at idle will be 12.5hz. This is smack in the middle of the frequency range those mounts were designed to have the GREATEST dynamic stiffness at.

QED

validius,

Thanks for posting this very interesting and useful Phd thesis. The problem in comparing to my observations is that I didn't notice the hydraulic mounts deflecting at all, between 500 and 1500 rpm. The thesis explains that the hydraulic mount works in both modes (coupled or decoupled) by deflecting. For all I could tell, they are just a solid block of rubber, and much stiffer than a M-B mount. Given their source (India), one might wonder what is truly inside.

I don't follow your discussion of the internal mass inside the mount. He even says on pg 35 that the mass of the mount's steel and fluid is negligible compared to the engine and car's body. The properties of the mounts that he measured (both hydraulic and solid rubber mounts), are their spring constant and damping, both of which vary with amplitude and frequency (non-linear). Don't intend to be smug, but that is exactly what I said.

All you guys keep talking about how the large vibrations at idle were due to these mounts deflecting too much - "OM617 ... easily exceed the IOD of the mount", "OM617 is heavier than an LS engine", etc. But, I didn't notice them deflecting at all!

As I read the thesis, the hydraulic mounts worked much better than the solid mounts around 10-15 Hz (pg 38, 1st paragraph, great at 9 Hz engine resonance). The concern was that they can be a bit worse at higher frequency (>150 Hz, same page bottom). Indeed, he says the same in the intro and all thru the paper. Indeed, there is nothing in the paper to suggest any difference in design criteria for a hydraulic mount for an I4, I5 or V8 engine application. validius might need to rethink his QED.

OK, there are a couple critical assumptions that need to be made here.

1. The mounts are built properly and are behaving as a hydromount should
2. The static stiffness of the mount is not so high that it alone could create the observed harshness

On to your reply:

With an IOD likely in the 0.3-0.6mm range peak-to-peak(p2p) you are not going to be able to perceive much deflection.

Remember, once you are at or above the resonance of the mount, the dynamic stiffness can be 2-3x+ the static stiffness, particularly at amplitudes of less than 1mm p2p (which is what it would need to be based on your observation).

The mass of the fluid inside the mount moves through tuned passageways to act as a mass damper. Outside of this scope, their mass is trivial compared to the rest of the system.

You are correct that the dynamic stiffness varies with amplitude and frequency however you originally conjectured that with increasing amplitude the dynamic stiffness would increase. This is not the case.

Once again, in the context of an IOD significantly less than 1mm and peak dynamic stiffness right around the input frequency at idle you are going to be faced with mounts operating at an extremely high dynamic stiffness, resulting in very small amplitude movements.

I applaud you for reading as far into that paper as you did. It's thick dry stuff to be sure.

The frequencies and amplitudes of the vibrations that engines create as they run varies widely not only within cylinder configurations but even more so between them. An i4 engine has a second order imbalance, resulting in a vibration at twice the frequency as the RPM of the engine. An i5 engine has a first order imbalance that causes the engine to want to rock back and fourth at the same frequency as the RPM of the engine. Mounts must be carefully tuned on an engine-by-engine basis to handle both smooth road shake, idle and many other metrics ( sometimes they even vary based on transmission because differing masses and CG have a meaningful effect on the response of the drive train ).

For more information on engine imbalance, watch the Engineering Explained series on the topic.

To visualize how a mass damper responds to differing input frequencies, check out this video: https://www.youtube.com/watch?v=Kus5nHW7Twc

Notice how the input force and the position become increasingly out of phase as the system approaches resonance. Also notice how the mass begins to stop moving as the input frequency exceeds the resonance of the system. This is what i referred to as 'locking up' in the previous post.


It is completely possible that the mounts are built so poorly that their static stiffness = their bulge stiffness at all times. Perhaps QED was a bit hasty however nothing you have observed precludes the possibility that the mounts you used were operating as designed.

I'm happy to answer questions or clarify anything you would like but you are not going to prove me wrong on hydromount theory. I used to design these things.

I found since that these Corvette mounts are made in a solid rubber version. Anchor adds "SR" to the PN to distinguish. Some Corvette guys wanted the solid rubber and were angry to find the ones in the "SR" box were hydraulic. They didn't say how they could tell. The ones I bought are Westar w/ same Anchor PN and no "SR", and were listed as "hydraulic" on rock. But, given the quality control in such India products, they might well be solid. I tried the classic "hard-boiled egg vs raw roll test", but couldn't tell if there is fluid inside (the fluid is very thick). They only cost $9 ea. Regardless, the 1" too high issue proved insurmountable. Even if I found truly hydraulic mounts, the engine brackets would have to be machined down at least 3/4" to fit. Perhaps another choice will show up someday.

In my case, the shaking was worse than I recalled after going back to factory mounts (w/ poly fill), but still much better than w/ the Corvette mounts. Turns out much was coming from a cracked Sanden bracket, so bad I removed it and found it cracked much worse than I thought (posted photos in Sanden bracket thread). Perhaps driving a few days with the Corvette brackets made it crack more. With an R4 compressor in, the idle seems OK, bothersome as always compared to most cars but apparently normal for these diesels.

Great effort Bill!

I am currently swapping a Ford Duratec engine into my BMW 2002 and I'm using hydraulic mounts from a BMW 318i. I haven't run the engine, but just shaking it by hand I can see the mounts deflect. I think this is still a viable avenue of research.