engine hard mounting vs thermal expansion 5

[ddace]

"ADMINISTRATION KILLS ENGINEERING"

Here it is

I am designing a 2.7Mw mobile genset. The engine is a MTU 20v4000 that has four(4) mounts that are attached to a skid base.

The installation specs require that mounting must be permissive to thermal expansion

1mm side to side with max resistance force of 25kN
2mm back to frt with max resistance force of 50kN

1. Would you agree that reaction forces due to thermal growth will depend on axial load on the mounts while taking in consideration friction factor on mounting surface.

2. Axial load is force of weight of engine on mount added to clamping force created by torque of fastener.

3. Assuming force of weight on each mount is pretty close to total weight/4 and that friction factor is the same at all 4 mounts, then the only variable left to work with is clamping force (i.e torque on fastener)If I set it up for 50 frt to back it will then be 50 side to side and vice versa at 20 for side to side.

4. Why do i get different specs for frt to back (50kn) than side to side (20kn). I have to set it up for 20kn at all four mount if i want to respect that maximum of 20 kn side to side (for both frt and back mounts) ?

5. If you could also confim the relationship between torque and clamping Force to be approx F=5T/D

6. any comments and/or general suggestions on this issue would be greatly appreciated.

Thanks

 

[CoryPad]

When you say hardmounting, do you mean a rigid solid (metal?) genset/mount with a through hole/slot that is connected to a rigid solid skid base? I don't see how this scheme will allow 1 to 2 mm movement unless the clamping force is small. If the clamping force is small, then forces like 25 to 50 kN will require large fasteners to prevent any axial movement. Large fasteners with small clamping force are prone to vibration loosening. Can you use a different attachment system, perhaps an elastomer mount that can provide direction-dependent spring rate and damping rate? The answer to question 5 above is that F = 5T/D is a rough approximation assuming the joint is rigid/seated, the fastener has some lubrication, and that no torque is used to overcome other effects (like deformed threads that are used for anti-rotation). I recommend that you perform a thorough analysis of the attachment system to prevent problems. You can review faq725-600 and faq725-536 for more information. Good luck.


Regards,

Cory

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[patprimmer]

I would suggest shoulder bolts and bearing pads, to obtain reliable results.

Then the bolt could be tight enough to resist vibration, and the joint could be loose enough to slide.

To obtain different travel and resistance in the different directions, the joint would need to be double decked, with one deck controlling one direction and another deck controlling the other.

I would much rather see a flexable (rubber) mount with rigid limits to movement.

Regards
pat

eng-tips, by professional engineers for professional engineers

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[mloew]

ddace,

The movement could be through elastic deflection of the mount stanchions or brackets and therefore the deflection and force prescribed will result in the maximum stiffness of the mounting structure being specified. The fastening method of the genset to the skid should be engineered for maximum resistance to vibration loosening.

Best regards,

Matthew Ian Loew
"Luck is the residue of design."
Branch Rickey

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[bradphillips]

You must also consider the reaction forces at the mounts due to motor torque and vibration.

 

[ddace]

Great input guys

Here are a little more details

yes it is a rigid mounting with a through hole/slot that is connected to a rigid solid skid base and i am using 1 inch grade 8, coarse tread, zink plated shoulder bolts and nuts have a nylon lock for vibration effect.

Flexible mounting is not an option because of flywheel housing attachement to alternator (no moment transfert possible due to engine torque)

the mount is a casting and i have machined some large washers (mild steel)to place between the mount and the skid (I guess that is what was ment by bearing pads ?) in order to allow the mount surface to slide along the washer: machined cast iron surface on machined mild steel

here is how i see this: Fmax = 25kN = (W+Fc)*Cf

W=weight on mount = engine tot/4 = 10070Kg/4*9.81=24.7KN
Fc= clamping force (KN)
Cf= friction coefficient ~ 0.24 (cast iron on steel ref: Gieck engineering formulas)

Fc=25-(.24*W)/.24= 79.5KN (max clamp force permitted)

Fc=5T/D , T = 79 500 N * 0.0254 m /5 = 404 N m

Now i intend to torque dry and would like to know if anyone has comments or observations.

Thanks to all

 

[bradphillips]

I noticed that the spec is for a maximum resistance force side to side and front to back. Is that common to specify for mounting? What is the intent of a max resistance force?

 

[mloew]

I still think it is a terrible idea to accommodate the deflection with letting the mount slide underneath the fasters. By forcing all of the deflection to be in the mount slipping, you are assuming that the mounts are not deflecting elastically. That must be a pretty stiff mount!

I am sure Cory will provide some additional commentary on the fastener calculations (single point value for friction and dry torquing and all). I do not see how intentionally allowing the joint to slip will provide any measure against self loosening. I think you should take another look at allowing the deflection to happen within the mounting structure, not the joint.


Best regards,

Matthew Ian Loew
"Luck is the residue of design."
Branch Rickey

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[patprimmer]

A shoulder bolt with a nyloc should not come loose if adequate in size, but the bearing surfaces will wear and distort, and the difference between static and dynamic friction will have an effect.

Also, this does not allow different resistance in the two planes.

The shoulder bolt and slot could work by having a horizontal hard mount with low friction but vertical bearing pads in both planes with rubber faces, with different hardness in each plane.

Have I explained this clearly? It sounds to me like I was talking with a mouth full of marbles.

Having said all that, I think Matthew's idea is the better aproach.

If you are worried about occasional excessive movement from inertia if Matthew's method is used, retaining straps can be added as a safety feature. An old hotrodders trick was to retain the original rubber engine and transmission mounts so as to retain comfort while cruising, but to have a chain between the chassis rail and the cylinder head, with just a bit of slack, so that on full power, the chain took the load before there was excessive distortion of the rubber mounts.

Just food for thought.



Regards
pat

eng-tips, by professional engineers for professional engineers

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[CoryPad]

As Matthew alluded, I do have criticisms. One data point (with two decimal precision) for friction coefficient is unfounded. If your design does not need two decimal accuracy, then assume a single decimal value and estimate high or low based on your needs (which are unclear - can you answer bradphillips' question?). If you need high accuracy, you need more data (friction coefficient scatter) and/or tests, etc.

Your use of a prevailing torque feature ("nylon lock" nut) invalidates the torque/tension equation you used, which already was of questionable accuracy. Again, more data and testing can help. I think patprimmer's first statement must be qualified in that loosening won't occur if adequate preload is achieved. Prevailing torque nuts have no effect on this, and neither does having a large screw diameter. Achieving proper screw pretension is the method to prevent loosening.

Lastly, I question how your equipment can have four hardmounts and deflect 1 to 2 mm in the joint plane. Can your equipment accomodate the strain imposed by 50 kN?

Regards,

Cory

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[desertfox]

Hi ddace

Apart from the variation of bolt torque particularly if done dry due to an uncertain friction coefficient which was mentioned in earlier posts, your assumption of each mount taking equal dead weight is only valid if the centre of gravity of the device is in the centre of the four mountings.
Further, are the mountings themselves subjected to the temperature rise ie;- I am thinking of the thermal stresses
generated vertically within the mountings and mounting
bolts/washers etc on which your planning the flexibility of
your system, if so the extra axial load due this would need
to be accounted for. Bradphillips also made a good point which was the dynamic loading on the mountings during operation which would invariably generate different loads in the mounting bolts.

regards desertfox

 

[ddace]

Again i repeat myself but this is grrrreat input

I will try to answer all questions

1.The purpose of max resistance force (as stated in the installation drawing of engine) is "To protect crankcase and engine mounts from excessive forces caused by thermal expansion." That is why I do not beleive that these mounts can take any deflection.The values i gave for resitance force and thermal expansion are specs given by manufacturer so i do beleive that the equipement can take it.

2. To design a somewhat sub mount between skid and engine mount that would be designed to allow the required deflection at a given force would probably require finit element analysis. That looks to me like killing a mosquito with a bazooka. This is not a rocket ship it is a deisel engine.

3. The reason why I did not try to allow different resistance in the two planes is to simplifie the problem by using the lowest criteria (25KN instead of 50KN )and to be safe.

4. The friction coefficient used was the highest of a range also again to be safe. I am not sure I understand the decimal issue " One data point (with two decimal precision) for friction coefficient is unfounded " ??? if the value i used does not seem good please advise of the correct value with references.

5. Where I would like some more details is the relationship between torque/tension equation and procedure. I would like to ensure I use proper torque procedure obtain correct preload on the joint. Please elaborate with references if at all possible.

Thanks again to all this is very helpfull

"ADMINISTRATION KILLS ENGINEERING"
ddace

 

[mloew]

Ddace,

I will repeat myself as well. I am not suggesting that the elastic deflection takes place on the mounts on the engine, but on the stanchions or other mounting structure on the skid. This will deflect elastically anyway, you mind-as-well design it to deflect the right amount in the presence of a thermal load. You may not need to perform Finite Element Analysis for this designed deflection mount. Use profiles with know cross-sectional properties and standard beam equations.

Good luck!

Best regards,

Matthew Ian Loew
"Luck is the residue of design."
Branch Rickey

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[tbuelna]

ddace:
an aluminum F1 engine, used as a stressed chassis member and mounted to the back of a composite tub, has the same thermal mismatch problem you mention. So, you may want to try the same solution that the F1 guys came up with: mount one side of the engine using "blade" mounts. The are stiff in two axes, but flexible in the third axis. Look at some pictures of some F1 engines from about 10 years ago. You'll see the "blade" type mounts tying the cam covers to the back of the tub.

regards,
Terry

 

[ddace]

What is a stanchion ?

 

[mloew]

From Merriam-Webster OnLine:

Pronunciation: 'stan-ch&n
Function: noun
Etymology: Middle English stanchon, from Middle French estanchon, from Old French, diminutive of estance stay, prop
1 : an upright bar, post, or support (as for a roof)

http://www.merriam-webster.com/cgi-bin/dictionary?stanchion

Please see the thread On-line Dictionary & Thesaurus(thread1010-87190) for useful links to on-line resources.

Best regards,

Matthew Ian Loew
"Luck is the residue of design."
Branch Rickey

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[GregLocock]

I think Matthew is on the right lines. In the automotive industry we try to mount our suspensions as rigidly as possible, and it takes a LOT of effort to achieve 25 kN/mm. I have never seen 50, when measured correctly.

Mounting the engine on short upright posts will allow you to fine tune the mounting point stiffness as the posts will put the main frame into local torsion, which is a notoriously difficult loadcase to design a stiff structure for. Int his case compliance is good, so that works in your favour.

However, I think you are wrong to dismiss elastomeric mounts, on at least two of the mounting points.



Cheers

Greg Locock

 

[ddace]

Well i guess everyone agree on one thing

My idea for allowing joint to slide is not popular...

Greg i like your observation and it nice to know that it is difficult to obtain a montage that is stiffer than the max permitted by specs. I will have to check but i think stiffness spec is the same frt to back and side to side 25kn/mm it's just that frt to back needs 2mm so that would explain the max force of 50kn. And side to side need only 1mm and 25kn.

thanks to all