3 cylinder diesel resonances, what's the best solution?

[David Mack]

Might it be worth considering solid mounting the engine?

[IanD]

Just now, David Mack said:

Might it be worth considering solid mounting the engine?

Only if you want horrible vibrations and resonances everywhere in the boat when the engine is running - solid mounting works for low-speed traditional engines, not for high-speed ones like this...

[Tracy D'arth]

I don't think you are going to stop this engine gyrating unless you get some weight on the mounting which as present is not loaded at all.

 

I would try adding a heavy weight to that corner of the engine and see what happens.

 

Think of a washing machine drum in the spin cycle. It has to speed up through many resonant frequencies  which vary as the weight in  the drum changes and the balance is changed. It survives because weights are added at critical points on the tub.

[IanD]

23 minutes ago, Tracy D'arth said:

I don't think you are going to stop this engine gyrating unless you get some weight on the mounting which as present is not loaded at all.

 

I would try adding a heavy weight to that corner of the engine and see what happens.

 

Think of a washing machine drum in the spin cycle. It has to speed up through many resonant frequencies  which vary as the weight in  the drum changes and the balance is changed. It survives because weights are added at critical points on the tub.

I don't know where you think the "many resonant frequencies" come from, there's usually one main one which is the drum (plus everything inside it e.g. wet washing) on the flexible supports -- weights are sometimes added to the suspended structure (not usually the drum) to add more mass and lower the resonant frequency, and there's often a big weight (e.g. concrete) in the bottom of the machine to add mass to this and try and stop it walking across the floor. As the machine spins up it goes through the main resonance (which changes with amount of washing, but not as much as you might think) and you can often see the whole machine shudder, but it rapidly goes through this as the drum speeds up so it's not a problem.

 

Adding more weight to the OPs engine will help even out the loads on the feet, but if the feet are each properly chosen for the load on them (e.g. 3mm compression) it won't alter the resonant frequency because you'd just need stiffer feet. The only way to lower the resonance to below idle is to have softer feet with more compression -- and *much* more, to halve the resonant frequency you need 4x the compression, which is why 3cyl engines have a bigger problem than 4cyl ones. I looked into all this when designing the inertia frame under my generator... 😉

Edited November 8, 2023 by IanD

[Tracy D'arth]

25 minutes ago, IanD said:

I don't know where you think the "many resonant frequencies" come from, there's usually one main one which is the drum (plus everything inside it e.g. wet washing) on the flexible supports -- weights are sometimes added to the suspended structure (not usually the drum) to add more mass and lower the resonant frequency, and there's often a big weight (e.g. concrete) in the bottom of the machine to add mass to this and try and stop it walking across the floor. As the machine spins up it goes through the main resonance (which changes with amount of washing, but not as much as you might think) and you can often see the whole machine shudder, but it rapidly goes through this as the drum speeds up so it's not a problem.

 

Adding more weight to the OPs engine will help even out the loads on the feet, but if the feet are each properly chosen for the load on them (e.g. 3mm compression) it won't alter the resonant frequency because you'd just need stiffer feet. The only way to lower the resonance to below idle is to have softer feet with more compression -- and *much* more, to halve the resonant frequency you need 4x the compression, which is why 3cyl engines have a bigger problem than 4cyl ones. I looked into all this when designing the inertia frame under my generator... 😉

But with one corner unsupported all calculations go out of the window!

[IanD]

39 minutes ago, Tracy D'arth said:

But with one corner unsupported all calculations go out of the window!

No they don't, that's not how it works -- if one corner isn't supported (no weight on it) then the foot isn't compressed but it still has the same stiffness, which is what affects the resonant frequency (all feet combined)

 

(actually not quite true because those rubber feet are nonlinear and get stiffer as you squash them, but the basic idea is correct)

 

How the engine moves when it rocks depends more on the inertia of the engine than the weight on each foot, but the stiffness of each foot affects how much each one moves by. A rocking motion (3cyl) is a lot more complicated to analyse than a bouncing one (4cyl), especially if the weights on the feet are different, but the same basic principle applies -- to get a low resonant frequency (below idle) you need soft feet, and much more so for a 3-cyl engine, which is why the foot manufacturers say these need special attention...

 

(and also why I made sure to use a 4-cylinder generator not the more common 3-cylinder ones)

Edited November 8, 2023 by IanD

[Ray T]

I don't know if Isuzu produce side rail bearers, but fitting a set cured resonance issues on our Lister Petter LPWS3.
RLWP of this parish fitted them.

[Peugeot 106]

Just a daft point. My engine beds were badly designed and not very rigid. The stern gland used to leak slightly more than it should. Stiffening the beds by adding small steel bars in situ cured this

[cheesegas]

On 08/11/2023 at 13:57, IanD said:

These Aquadrive mounts are soft and will easily fit in 65mm height:

 

https://www.peachment.co.uk/engine-mountings-couplings/

 

"It is essential to select the correct rubber grade and mount for the engine or generator in question. For reference, the chart above shows the level of deflection
(or squash) that will occur on each model of mount for a given loading.
In general you should aim for 3mm on the 50210, 4mm on the 50220, 5mm on the 50230 and 6mm on the 50240. The numbering system is simple: A 50223
mount is the 50220 model with number three rubber hardness and so on. Some engines, particularly those with less than four cylinders, require particular care and
your distributor will help with these."

Thanks, I'll give them a go. I think they may be designed to be used with an Aquadrive system however, which means the mounts aren't designed to take any thrust forces as the boat is effectively pushed along by the thrust bearing bolted to the hull rather than by the engine mounts in an installation like mine.

 

On 08/11/2023 at 14:40, Peugeot 106 said:

Have you sent this to R&D technical. My experience with them is that they can be abrupt but this us after all their bread and butter. Their opinion could be helpful even if their mounts aren’t cheap and they probably have direct experience of your problem

you have clearly explained the problem so they should have an answer

I called and then followed up with an email when I ordered the first set of mounts but no one replies - I'll try again though. I've got a feeling they may ask for an accurate weight on each mount which I can't give though...might have to rig something up with a borrowed chainblock and strain gauge from work to get a better idea.

 

21 hours ago, Ray T said:

I don't know if Isuzu produce side rail bearers, but fitting a set cured resonance issues on our Lister Petter LPWS3.
RLWP of this parish fitted them.

Good idea but almost definitely not - this is an industrial engine that wasn't marinised by anyone like what Beta do to Kubotas. 

 

7 minutes ago, Peugeot 106 said:

Just a daft point. My engine beds were badly designed and not very rigid. The stern gland used to leak slightly more than it should. Stiffening the beds by adding small steel bars in situ cured this

The beds are made pretty strongly in my boat - the engine is moving on the rubber mounts causing the gland to leak, I think the elasticity of the steel beds is negligible in this.

[BWM]

You definitely need to use the manufacturers spec, or something with identical properties. Our last boat had a 3KR1 and it was extremely smooth, right through the rev range, but when the originals failed I was supplied firmer ones by a well meaning friend and you could almost feel your eyes shaking in your head! This was quickly cured by fitting the correct ones. 

 The very short shaft is probably exacerbating the vibration too.

 

[IanD]

1 hour ago, BWM said:

You definitely need to use the manufacturers spec, or something with identical properties. Our last boat had a 3KR1 and it was extremely smooth, right through the rev range, but when the originals failed I was supplied firmer ones by a well meaning friend and you could almost feel your eyes shaking in your head! This was quickly cured by fitting the correct ones. 

 The very short shaft is probably exacerbating the vibration too.

 

 

The best option by far, if such can be found (good luck with that...) and they will fit in the available height... 😉

 

But note that recommended tickover is 1000rpm and the OP's engine is using 800rpm, which could very possibly get back into the danger zone if the OEM mounts are designed to put the resonance not far below 1000rpm... 😞

Edited November 9, 2023 by IanD

[BWM]

32 minutes ago, IanD said:

 

The best option by far, if such can be found (good luck with that...) and they will fit in the available height... 😉

 

But note that recommended tickover is 1000rpm and the OP's engine is using 800rpm, which could very possibly get back into the danger zone if the OEM mounts are designed to put the resonance not far below 1000rpm... 😞

I never had a tachometer, so set tick over to the point where it sounded right. 

[Peugeot 106]

Suggest you make a plan view of the mount configuration, engine gearbox, Shaft , coupling and stern gland.

 

add to it weights from manufacturers info for engine and gearbox, coupling and shaft Add rough position of flywheel. ( you can consider engines as homogenous mass but the flywheel weight and position needs to be taken into account) They should have rough idea of flywheel   weight but you can estimate how far back it is on the engine. 

 

Roughly someone needs to calculate weight distribution over from and rear mounts. Not rocket science R&D must do it all the time. Then it’s a case of whether it’s a modern “ fast “ diesel or old slow one eg 1500 rpm ( engine not propspeed) and whether piston stroke is long (old) or short(modern)

 

R&D should be able to calculate near enough from this. No need to unbolt everything and weigh. I assume it is a modern fast revving (around 3000rpm)

 

suggest if you can’t get hold of R&D in UK try sending to office in USA. R&D aren’t brilliant at answering the phone or explain there answers but they have been at it for a long time.

 

good luck

[cheesegas]

11 hours ago, BWM said:

You definitely need to use the manufacturers spec, or something with identical properties. Our last boat had a 3KR1 and it was extremely smooth, right through the rev range, but when the originals failed I was supplied firmer ones by a well meaning friend and you could almost feel your eyes shaking in your head! This was quickly cured by fitting the correct ones. 

 The very short shaft is probably exacerbating the vibration too.

The shaft seems long compared to most trad stern boats without an engine room I've looked at - also, the vibrations are more without the propshaft connected as I expect the shaft is anchoring it to the hull in some respect. Probably a good 2' (realised I put 1' in the original post, typo sorry!) between output flange and gland.

 

Unfortunately the manufacturer is Isuzu who at this time made industrial engines for any application, so the complete engine and gearbox package was never installed by a mariniser like Beta. I've checked with them and they never used this block in their early day. This all means that there's no recommended manufacturer's spec mounts for this engine/gearbox.

 

10 hours ago, IanD said:

The best option by far, if such can be found (good luck with that...) and they will fit in the available height... 😉

 

But note that recommended tickover is 1000rpm and the OP's engine is using 800rpm, which could very possibly get back into the danger zone if the OEM mounts are designed to put the resonance not far below 1000rpm... 😞

Sorry, think I may have confused things by mentioning so many speeds! When I first got the boat the engine was set to 800rpm, it's now at 1000rpm as per the service manual. When in gear at tickover it drops to 950rpm which is also smooth.

 

10 hours ago, BWM said:

I never had a tachometer, so set tick over to the point where it sounded right. 

The engine is perfectly smooth, barely vibrates and sounds right at a 1000rpm tickover, it's just the rev range between that and 1300rpm is the issue. Setting tickover to 1300rpm sounds a tad fast but is smooth, but the gearbox doesn't sound very happy though and it makes mooring difficult.

 

7 hours ago, Peugeot 106 said:

Suggest you make a plan view of the mount configuration, engine gearbox, Shaft , coupling and stern gland.

 

add to it weights from manufacturers info for engine and gearbox, coupling and shaft Add rough position of flywheel. ( you can consider engines as homogenous mass but the flywheel weight and position needs to be taken into account) They should have rough idea of flywheel   weight but you can estimate how far back it is on the engine. 

 

Roughly someone needs to calculate weight distribution over from and rear mounts. Not rocket science R&D must do it all the time. Then it’s a case of whether it’s a modern “ fast “ diesel or old slow one eg 1500 rpm ( engine not propspeed) and whether piston stroke is long (old) or short(modern)

 

R&D should be able to calculate near enough from this. No need to unbolt everything and weigh. I assume it is a modern fast revving (around 3000rpm)

 

suggest if you can’t get hold of R&D in UK try sending to office in USA. R&D aren’t brilliant at answering the phone or explain there answers but they have been at it for a long time.

 

good luck

Thanks, I'll have a go at this when I'm back! Yep, it's a 1000cc three cylinder high speed engine. The difficulty is that the side of the engine with the alternator/manifolds/starter is a lot heavier which I think will be difficult to estimate without measuring weights; it's more unevenly distributed left to right than back and front. I'll have a go at doing it by measuring deflection and if that gives some nonsense answer, I might need to weigh it.

[Peugeot 106]

I think you need to get hold of R&D and get some advice. I don’t know how much side to side weight difference matters. I thought it was about front foot and back foot distribution. They didn’t ask me if the engine had all it’s ancillaries on one side. Suggest you keep ringing or try USA.

They don’t need weights to the nearest gram. You can make a very good estimate off the published figures or they may know the engine but you do need an accurate footprint of the feet and configurations of components as overhangs need to be factored in

There may be no weight on one foot because the mountings aren’t correctly adjusted and it is in effect sitting on three feet like a milking school. I suggest if you haven’t read about it you read up about it there is loads written about how to align it all but it’s still a very frustrating if you are a beginner. Definitely a job where you need to be calm with plenty of time lock out  the cat, kids,wife, concubine.

All Best

[cheesegas]

12 minutes ago, Peugeot 106 said:

I think you need to get hold of R&D and get some advice. I don’t know how much side to side weight difference matters. I thought it was about front foot and back foot distribution. They didn’t ask me if the engine had all it’s ancillaries on one side. Suggest you keep ringing or try USA.

They don’t need weights to the nearest gram. You can make a very good estimate off the published figures or they may know the engine but you do need an accurate footprint of the feet and configurations of components as overhangs need to be factored in

There may be no weight on one foot because the mountings aren’t correctly adjusted and it is in effect sitting on three feet like a milking school. I suggest if you haven’t read about it you read up about it there is loads written about how to align it all but it’s still a very frustrating if you are a beginner. Definitely a job where you need to be calm with plenty of time lock out  the cat, kids,wife, concubine.

All Best

Thanks. The engine is correctly aligned and adjusted - I've measured with calipers to ensure that every leg is exactly 57mm between leg and bearer, and the propshaft alignment is correct. I'm not a mechanic but the alignment is very close to perfect, I've been removing the Centaflex and using a solid coupling to get the alignment right with feeler gauges. 

 

When the engine is perfectly flat in this configuration, the back right leg takes very little weight, whilst the back left is almost at full deflection. 

Edited November 10, 2023 by cheesegas

[Tony Brooks]

15 minutes ago, cheesegas said:

When the engine is perfectly flat in this configuration, the back right leg takes very little weight, whilst the back left is almost at full deflection. 

 

That suggests that the engine is not evenly supported on the diagonal and we know that tends to result in the engine rocking. No advice of help, I am sorry to say.

[Peugeot 106]

13 minutes ago, Tony Brooks said:

 

That suggests that the engine is not evenly supported on the diagonal and we know that tends to result in the engine rocking. No advice of help, I am sorry to say.

Agreed. That’s what I have read. It’s not difficult to get most engines balanced on three feet which causes excessive rocking. These is plenty written about it. R&D are 01462 892391. Have you tried them again?

 

 

 

 

 

[cheesegas]

4 minutes ago, Peugeot 106 said:

Agreed. That’s what I have read. It’s not difficult to get most engines balanced on three feet which causes excessive rocking. These is plenty written about it. R&D are 01462 892391. Have you tried them again?

Thanks. Unfortunately I’m out the country until mid Dec, working out in the UAE where it’s very difficult to make voice calls to landlines. I’ve emailed and will call in Dec and then update this thread!

[IanD]

14 minutes ago, Peugeot 106 said:

Agreed. That’s what I have read. It’s not difficult to get most engines balanced on three feet which causes excessive rocking. These is plenty written about it. R&D are 01462 892391. Have you tried them again?

 

 

3-cylinder engines suffer from fore-and-aft rocking regardless of how many feet they've got, it's caused by an unbalanced rotating couple at the engine frequency which is hard to deal with at low rpm by using flexible feet -- in comparison a 4-cylinder has a vertical imbalance at twice the rotation frequency which is much easier to deal with. In the OPs boat the prop shaft will try and prevent this rocking happening, but the result will be huge sideways bending forces on the couplings and stern bearing if the resonance happens which is likely to cause damage.

 

Adding opposing counterweights on flywheel and pulleys (needs doing at both front and back of engine) cancels out the fore-and-aft rocking but results in a side-to-side rocking instead which is just as bad or worse -- the only way to fix this is with a counter-rotating balance shaft if these crank weights are added, or two counter-rotating balance shafts if the crank is balanced, but few engines do this and certainly not this one. The other solutions used in cars -- extremely soft or hydraulic engine mounts -- can't be used here because they won't fit, and also have to withstand propeller thrust.

 

If you can get this rocking engine/foot resonance below the idle frequency (1000rpm = 17Hz) then the problem will go away, but this needs *all* the feet to be soft so they have fairly large deflection under static load (several mm) -- this is what Yanmar do in their 3-cyl marine engines. If the weight distribution is uneven across the feet then the individual foot stiffnesses need to be varied to get roughly the same static deflection on each, which is what the OP is proposing to do. Usually this is just needed front-to-rear, but it looks like the OP will have to do this side-to-side as well.

 

I expect this will all be confirmed by R&D (if they give any response), it's basic mechanical engineering... 😉

[Peugeot 106]

When I had classes in vibration regrettably I just closed my eyes and ears and learnt the formulae to pass the exam! But there must be people who payed more attention like you as there are loads of 3 cylinder engines in all sorts of boats working perfectly satisfactorily. My only dalliance was with 2 cylinders and I’m no expert

[IanD]

25 minutes ago, Peugeot 106 said:

When I had classes in vibration regrettably I just closed my eyes and ears and learnt the formulae to pass the exam! But there must be people who payed more attention like you as there are loads of 3 cylinder engines in all sorts of boats working perfectly satisfactorily. My only dalliance was with 2 cylinders and I’m no expert

 

I investigate this issue (engine/feet resonances) in detail when designing the inertia frame my generator sits on, including simulating all the resonances to make sure it didn't hit any -- so I'm pretty sure what I'm saying can be backed up 🙂

 

Loads of 3-cyls -- yes there are, and I'll put money on them having softer mounting feet than the OP, because that's what the engine manufacturers always recommend... 😉

 

(or they idle below any resonance and cruise above it and don't spend any time at the "bad" rpm)

 

This is all for high-speed engines like the OPs -- low-speed trad engines are rigidly mounted to engine beds, obviously... 😉

Edited November 10, 2023 by IanD

[hughc]

When we fitted a Ford New Holland BSD3 to Wyrd in 1989 it sat on R&D mountings. This engine is over three litres and produces  42 HP @ 2200RPM. It has an horrific nodding couple characteristic of three cylinder engines. Very soon after fitting it broke  the 5/8" bolts on two of the couplings. R&D came out and suggested that we fitted much heavier mountings and a very heavy crankshaft pulley with a rubber insert to try and counteract the out of balance forces. The engine destroyed two of the mountings in short order. Ford New Holland suggested that three cylinder engines are inherently unsuitable for the flexible mounts sold for small marine use. C-power who provided the engine produced a pair of full length steel beds to fit. The engine was then mounted on hardwood beds which were bolted to the orginal steel beds with a 1/2" insulating layer of Tico composite from James Walker and Co. This totally isolates the engine from the beds as the bolts are encased in Tico tubes with washers. The Tico absorbes the higher frequency harmonics. The beds take care of much of the lower frequencies. The pulley reduces the out of balance forces. In nearly thirty five years I have never had to tighten the mounting bolts and Wyrd has spent a fair bit of time towing. The engine runs at 1000 to 1200RPM on the cut and up to 1500RPM on the river. It is worth noting that when Lister marinised this engine as the the CRK3 they had it 'walking across the yard' when mounted on temporary beds.

 

[cuthound]

In answer to the question posed in the thread title, surely the answer is to replace  the engine with one with an even number of cylinders... 😅😂

[IanD]

58 minutes ago, hughc said:

When we fitted a Ford New Holland BSD3 to Wyrd in 1989 it sat on R&D mountings. This engine is over three litres and produces  42 HP @ 2200RPM. It has an horrific nodding couple characteristic of three cylinder engines. Very soon after fitting it broke  the 5/8" bolts on two of the couplings. R&D came out and suggested that we fitted much heavier mountings and a very heavy crankshaft pulley with a rubber insert to try and counteract the out of balance forces. The engine destroyed two of the mountings in short order. Ford New Holland suggested that three cylinder engines are inherently unsuitable for the flexible mounts sold for small marine use. C-power who provided the engine produced a pair of full length steel beds to fit. The engine was then mounted on hardwood beds which were bolted to the orginal steel beds with a 1/2" insulating layer of Tico composite from James Walker and Co. This totally isolates the engine from the beds as the bolts are encased in Tico tubes with washers. The Tico absorbes the higher frequency harmonics. The beds take care of much of the lower frequencies. The pulley reduces the out of balance forces. In nearly thirty five years I have never had to tighten the mounting bolts and Wyrd has spent a fair bit of time towing. The engine runs at 1000 to 1200RPM on the cut and up to 1500RPM on the river. It is worth noting that when Lister marinised this engine as the the CRK3 they had it 'walking across the yard' when mounted on temporary beds.

 

 

As Ford suggested, a 3l three-cylinder is certainly going to be a nightmare for flexible feet, especially because the idle speed is presumably well below 1000rpm -- the OPs engine is a third the capacity so less difficult (but still not easy).

 

I'd be very surprised if the Tico did anything for the out-of-balance forces -- even harmonics of them -- because it's *much* too stiff, but it will help with the high-frequency compression ignition noise (diesel knock) and stop this spreading through the hull. The 3-cyl rocking is stopped by the rigid engine beds connected to the hull baseplate and sides, the sheer mass and rigidity of this stops the engine moving -- but all the energy goes into the hull so it may be noisy if the frequency happens to coincide with any hull or cabin resonances at certain rpm, I've certainly noticed this on boats I've been on with rigid-mounted engines. But it's still the best solution for engines like a 3-cyl which have poor balance...

 

4 minutes ago, cuthound said:

In answer to the question posed in the thread title, surely the answer is to replace  the engine with one with an even number of cylinders... 😅😂

 

That would certainly help, 4 is a good choice. Maybe that's why 4-cyl engines are so popular? 😉

Edited November 10, 2023 by IanD