Addressing MEP-802a vibration

[Digger556]

I expect to get mixed replies on this one, but I've been suffering from Engineer Brain lately and I want to take a swipe at reducing the chassis vibration of an 802a.

Background: I've own an 802a since 2018 and an 803a since 2023. I really like how smooth the 803a is, but most situations call for the 802a. I came across a parts bucket of 802a and got it running, but the vibration is the worst I've seen, yet it still feels somewhat normal for these machines. My concern is the longterm damage to wires, gauges, relays, etc.

This has led to a nagging desire to re-analyze the approach taken by the OEM engineers to isolate the vibration-prone two-cylinder engine and see if there is room for improvement. I've researched the general approach to driven vibration isolation and will start by analyzing the natural frequency of the DN2M's isolators and compare that to the driven frequency, calculating the transmissibility of the system. From there, make a judgement call on whether there is room for improvement by modifying the isolator rate or damping coefficient. If meaningful gains could be realized, I will experiment. If not, I will explore the level-of-effort involved in adding an active balancing element to the engine.

Feel free to follow along, comment, or ridicule. This felt like a fun learning exercise.

 

[m38inmaine]

I like the idea, standing by.

 

[WWRD99]

Well just in my automotive world I have seen many NVH things attached randomly on suspension parts and engines over the years. It is usually a bracket with some sort of heavy rubber mounted to a strut or engine mount. I am sure the engineers spent time finding the correct weight to counter the vibration of whatever part they found in that acoustic range. I am not sure exactly how they do it but cars are much quieter since. Could possibly try out a few different materials mounted on the engine and see if anything changes. I think it comes down to the tiny motor mounts they have too. Fluid filled ones, I think, would do a lot more. Mazda uses one on its 3 engines side motor mount that might be able to adapt or a honda front mount. would have to make them fit though. They both are just round mounts with a top bolt...but much bigger. I think the honda one would be easier to fit. This is a pic of one...would have to modify the floor in the case lower to fit it and make a top bracket to bolt to the motor. These are vacuum operated too so it could be pumped up to hold more vibration.

 

[Light in the Dark]

Have you confirmed you have the right isolators in place on the machine in question? 803 looks the same, but different molded number and different material characteristics.

 

[2Pbfeet]

I have chased NVM on more than an few items, and speaking as a non-professional, it always feels to me like a swamp. Folks who make zillions of something can put it into a computer model and simulate it or test it and find the combination of tweaks and absorbers to make a difference. For the rest of us who have time and material budgets...

@Digger556 This is for entertainment, right? There is no thought of a return on investment for your time or money, right? Do you want this unit fixed, trailer mounted, or mobile?

Given that you wrote that this unit was a "parts bucket", I would first double check the injection system to make sure the pumps and injectors are matched and working identically, (I would start by checking the exhaust manifold to see if the exhaust temperatures from each cylinder match after running under load for a bit), and then I'd be looking at how better to bolt it down to a large block of concrete with appropriately sized vibration dampeners. Then, if it were me, I agree with @Light in the Dark, I would look at changing the motor mounts, (rubber ages, and I have found that cracks aren't always obvious). After that, I might add some mass weighted vinyl to the enclosure, but that's not cheap. Then if you are going all out, I would pull the crankshaft and pistons and balance all of them ($$$). I hear the siren call of the engineer brain on the active balancing system, but gosh, that seems like an undertaking. To me, I think designing and building a second dog house to enclose the whole 802 unit would be easier and more likely achieve the reduction in vibration and noise.

All the best,

2PbFeet

 

[Digger556]

Have you confirmed you have the right isolators in place on the machine in question? 803 looks the same, but different molded number and different material characteristics.

Unfortunately yes. Of the three running 802a's, I'm not sure why this one particular unit shakes more, but I still need to verify the injectors are functioning correctly (pop test) and play with the timing slightly to see if that makes a difference. My nicest, newest unit is a 200 hr machine in perfect condition and it's level of vibration is in the middle of the other two. The oldest unit with 200-ish hours and stuck rings vibrates the least. They all shake more an desired, but it's an 802a.

 

[Digger556]

I have chased NVM on more than an few items, and speaking as a non-professional, it always feels to me like a swamp. Folks who make zillions of something can put it into a computer model and simulate it or test it and find the combination of tweaks and absorbers to make a difference. For the rest of us who have time and material budgets...

@Digger556 This is for entertainment, right? There is no thought of a return on investment for your time or money, right? Do you want this unit fixed, trailer mounted, or mobile?

Given that you wrote that this unit was a "parts bucket", I would first double check the injection system to make sure the pumps and injectors are matched and working identically, (I would start by checking the exhaust manifold to see if the exhaust temperatures from each cylinder match after running under load for a bit), and then I'd be looking at how better to bolt it down to a large block of concrete with appropriately sized vibration dampeners. Then, if it were me, I agree with @Light in the Dark, I would look at changing the motor mounts, (rubber ages, and I have found that cracks aren't always obvious). After that, I might add some mass weighted vinyl to the enclosure, but that's not cheap. Then if you are going all out, I would pull the crankshaft and pistons and balance all of them ($$$). I hear the siren call of the engineer brain on the active balancing system, but gosh, that seems like an undertaking. To me, I think designing and building a second dog house to enclose the whole 802 unit would be easier and more likely achieve the reduction in vibration and noise.

All the best,

2PbFeet

This is definitely a type II fun project with no expectation of ROI. I like the 802's, but don't accept the level of vibration, even out of good running units. If I could spec an alternate isolator that resulted in 25% reduction in vibration, I would consider that a win. If I could get 50% or more reduction with a combo of isolators and mass dampers, that would be awesome. I may get no where, but figure it's worth a try, as I have 3 units to play with.

Alternately, slapping 1000 lbs of mass to the frame would make a significant improvement too.

 

[2Pbfeet]

Unfortunately yes. Of the three running 802a's, I'm not sure why this one particular unit shakes more, but I still need to verify the injectors are functioning correctly (pop test) and play with the timing slightly to see if that makes a difference. My nicest, newest unit is a 200 hr machine in perfect condition and it's level of vibration is in the middle of the other two. The oldest unit with 200-ish hours and stuck rings vibrates the least. They all shake more an desired, but it's an 802a.

Well, with three, you have lots of room to compare and contrast. Sounds like a fun project.

I think starting with the pop test on the injectors is a great place to start. After truly imbalanced crank/piston/rods, I think imbalanced injection is the one most likely to add excessive vibration. However, with an engine that had been through disassembly and reassembly by different folks, I think there are lots of possibilities. How is the compression?

All the best,

2Pbfeet

 

[Digger556]

Well, with three, you have lots of room to compare and contrast. Sounds like a fun project.

I think starting with the pop test on the injectors is a great place to start. After truly imbalanced crank/piston/rods, I think imbalanced injection is the one most likely to add excessive vibration. However, with an engine that had been through disassembly and reassembly by different folks, I think there are lots of possibilities. How is the compression?

All the best,

2Pbfeet

Havent run a compression test, but it pops off and seems to run strong under load. I was only able to load test it in short bursts because I kept getting an overload shutdown alarm. I have since use Deoxit on the mode switch and need to retest. This unit sat in Missouri for a long time in high humidity.

 

[Guyfang]

LITD wrote:
Have you confirmed you have the right isolators in place on the machine in question? 803 looks the same, but different molded number and different material characteristics.
Do not ignore this post. I saw it happen many, many times.








Item #20 above, can cause you problems also.
Item #16 & Item #32 and Item #39, if loose, could cause problems.

 

[rickf]

In addition to what Guy says about the loose bolts, if not torqued evenly that can also cause vibration. I have seen that several times on automotive flywheels. Bolts were tightened by hand an no torque wrench was used. Once I loosened all the bolts and torqued them in the proper sequence it was smooth as can be.

 

[Digger556]

LITD wrote:
Have you confirmed you have the right isolators in place on the machine in question? 803 looks the same, but different molded number and different material characteristics.
Do not ignore this post. I saw it happen many, many times.






View attachment 933579

Item #20 above, can cause you problems also.
Item #16 & Item #32 and Item #39, if loose, could cause problems.

Copy that Guyfang!

I pulled this thing apart last year and measured considerable runout at the end of the rotor shaft. Some was flexing of the [35] plate, but I fully disassembled the whole rotor, adapter plate and flywheel, measuring excessive runout on each interface. I've designed and machined crank adapters in the past for custom engine applications and typically aim for less than .0005" runout, so seeing .009" TIR on the end of rotor was shocking.

I spun up all the parts in the mill or lathe and squared all the interfaces, then bolted the adapter plate to the flywheel and squared them as a matched unit, then installed a new bearing and torqued all the bolts to spec per the TM. Crank to flywheel runout was <.0005" and the end of rotor was now .001".


After reassembly, vibration was noticeably reduced, but still was shaking the control panel a bit more than normal. Taking some high speed video of the engine, the isolator displacement looked normal, so it's still a bit of a mystery. I did confirm I have the right isolators by P/N, checking the TM and other 802s.

 

[2Pbfeet]

And the prize for attention to detail goes to...

Wow! Nice job.

All the best,

2Pbfeet

 

[Toolslinger]

That's some nice work!
Did you think about running it through a balancing shop?

 

[rickf]

I was thinking that all of that glue and sealant in there could not be even close to balanced.

 

[2Pbfeet]

It would have been pretty straightforward to test the balance with a second bearing support for the engine side. I don't see any balance drill holes in the fan assembly, which argues to me that there wasn't a factory process, but doesn't say anything to me about potential imbalance. The wiring and laminations are very heavy compared to sealant mass.

All the best,

2Pbfeet

 

[Digger556]

That's some nice work!
Did you think about running it through a balancing shop?

In retrospect, I wish I would have while it was apart. The runout was so bad, I thought for sure that was the issue.