Mercedes Engine - ST Generator Mounting

[WStayton]

One of the issues that Tom Osborne brought up in response to myriad questions has posed a new question about holding things down. 

My Mercedes Engine has "feet" on it that are the result of its previous life as Marine engine.  Each of the four feet has approximately at 0.50 inch hole in it base flang.  I don't have the exact dimension right at hand, but I THINK that they are something like 24 inches apart side-to-side. though the two forward ones are about 1.5 inches shorter than the two rear ones.

A 24 kw ST Generator-Head has 19 mm holes, which is just a shade less than three-quarters of an inch (0.748031496 to be exact! <grin>).  These holes are 31.9 cm, center-to-center, side-to-side.  This is about 12.520 inches.

The Mercedes Transmission that I envision putting between them and plugging directly into the Flex-plate that is currently under the Marine Transmission.  As luck would have it, it appears that the input shaft to both the marine transmission and the road transmission are identical in terms of diameter and splines.  The road transmission has a mounting that is normally secured to the vehicle, and has two bolt holes an either side of the output shaft which are about 0.050 in diameter - that is eyeball, I haven't yet mic'ed them.

I'm planning to build a base frame out of three layers of 2" x 6" nominal softwood stock, so the overall assembly will be 4.50 inches thick with each "rail" being 5.50 inches wide.  I plan to run two of these for and aft, under the whole installation and have the top layers be one piece, and the center layer made up of pieces, most of the for and aft, but two cross wise under the engine, maintaing them so that they are 24" on center, so the engine mounts are centered on them, side to side.

Since the generator is much narrower, I will have to have two additional 4.5" x 5.5" "runners" under it that would be inboard of the two runner from the engine, with the generator runners being 12.5" on centers under the generator and two cross pieces in the middle layer to hold the four total runners firmly in "register" so that they don't spread.

Everything will have a trough carriage bolts from the bottom through the runner and then through and shim blocks to make everything th right height and the smallest ones would be the ones through the engine "feet" that would be half-inch.

I also envision a cross piece under the transmission the give me something to bolt that transmission to, again with wood shims to height.

Between the generator and the transmission will be a lovejoy coupling, completely unsupported, other than by the transmission output shaft and the generator input shaft.

My question is:  Is this whole structure sufficiently stiff and robust for what I am asking it to hold?

One thing I have thought about was should I have some sort a large washer, or maybe a piece of .250 inch thick steel stock running along on top of the rails, between each engine leg or generator bolt to try to alleviate the problem of having the wood get "squashed" and thus bring everything out of alignment.

Also, what torque figure do you think I should use for the various hold-down bolts?  Keep in mind, it can't be too much torque or I will "squash" the wood before I even get it going and make it impossible to get everything lined up!

Side thought:  I suppose I could use some amount of "squashing" to perform that final alignment; get is as close as possible with practically no torque on the bolts and then torque/"squash" to bring everything in to alignment.

As I envision the process, I will have to torque down the engine/trasnmission as one piece and then check the the centerline of the front of the crankshaft and the back of the transmission out put shaft are at the right/same height and then check that both ends of the generator shaft are at the same height, as each other and as the engine/transmission, and then install the Lovejoy coupling, since it is floating and only cares if the shafts are aligned, not what the absolute height is.

I await your criticisms, comments, etc.

Regardz,

Wayne Stayton

[Crofter]

  In my opinion wood is not stable enough to maintain alignment for best coupling life. Any torsional flexing of the assembly also will effectively change relative height of shafts at the joint. Laminated is not as rigid as solid timber unless glued and pressed. Seasonal changes from moisture, different grain structure, temperature fluctuations, vibration etc., will continue to "work" on the load bearing areas of the wood.

  The life of the spider is greatly reduced by misalignment. Couplings should not be forced into doing the work of a pair of universal joints. Decking the whole base with a piece of 1/2 steel plate would help a lot.

[mike90045]

Oak would be a much better choice than softwood.  Needs to be really stiff and strong, not just some 2x4s flopping around.  Steel I  or H beams wood be better!

[WStayton]

There are a few things I left out . . . yes, even in a 5000 word expose, I forgot a few things!  <grin>

The support as planned, would be fabricated of 1.5 " x 5.5" clear, or possibly number 2, construction pine as readily available nearly everywhere - think Lowes, 84 Lumber, etc., etc.

These pieces of wood would be glued and screwed together.  The glue would be resourcinal (sp?) as is commonly used for laminating marine timbers and has excellent holding properties and is waterproof, even if it is a pain to work with.  Strength of the joint nowmally exceed the strength of the wood it is fabricated between.  Screw would be 2.25" between layer one and layer two, so as not to protrude on the bottom, and 4.25" on any subsequent layers, to ensure that the top layer was held down to at least two layers under it.  The screws would be put in in a 2 - 1 - 1 pattern where there are two screws across the width, spaced 1.25 " either side of the centerline and the sindle screw would be on the center line.  The distance between the 2 screws and the single screw would be 3 inches along the lenth of the timber.  These screw are only really to clamp the whole mess together until it drys/cures.  Ecah layer would probably have to be layed up separately, since the curing time of the resourcinal is suffciently short that unless you did it in a walk in cooler, the batch you mixed up for layer one would be at least partially set by the time you finished screwing layer one to layer two.

About the .250 inch thick 5.50 wide steel on top:  This would be used as the last layer on top of any wood, probably bedded in marine epoxy, just to make sure it didn't wiggle till everything was tightened down.  There is a problem with making it one piece whole length, in that there are sever diffenerent height's required:  two heights for the different length "feet" on the engine, a different height for the gear-box and a different height, still, for the generator.  My though was to make any piece of the .250 inch steel at least 5.50 inches long, so that every hold down would be spread over a minimu of 30.25 square inches of steel.

The whole mount assembly would sit on a five inch, nominal, concrete slab.  I was thinking to hold it down with a total of ten 0.50" x 10"-12" (depending on where they went trough the mounting and therefore the munting thickness at that point) lag screws, screwed into a lead anchor.  I have had good luck setting the lead anchors in epoxy but that was for a much less repetetive load (large tent holdowns).

mike90045:

I selected common pine due to its common availability - oak is much scarcer in 1.5" thicknesses and due to cost.  Common pine is going to be somewhere in the neighborhood of $1 per board foot, while oak is going to be more like $3 per board foot.  Since the project requires about 76 board feet of 2" x 6", I'm reluctant to spend $150 dollars that isn't really required.

Construction timber is generally felt, by most civil engineers, to have a compressive strength of 6400 psi while oak is something in the range of twice that.  My 5.50" x 5.50" x .250 square mountings, minimum, would thus have a compressive strength of about 40,000 pounds each in common pine which, I should think, is enough, even if I have moderate torque on the hold down bolts.

Crofter:

I realize that the lovejoy is VERY sensative to alignment, but I am hoping that the foregoing is adequate to insure that the alignment will be sufficently accurate to satisfy the lovejoy. I have though that the most critical part of securing alignment is going to be the metal shims that will be required on top of the wood and .250 steel pile to get exact alignment


Otherwise, I go to 6" thickwall box tubing and weld-up everything - its just that I am more experienced with glueing-and-screwing than I am with welding, on a do it yourself basis.  Worst case is that I would draft my brother, who is a VERY good welder, to do the welding in his shop, but he has little patience with fiddley projects so that really is a last resort.

With the forgoing used and in place, is it adequate?

[potter]

Boat engines are commonly bedded on wood with little problems if done properly. also with alignment issues.

   Potter

[Crofter]

The calculations on total load bearing ability of the proposed 1/4" steel plates seems to make the assumption that they will not bend but rather distribute load evenly. Divide and conquer! 3/4" thick would be more likely. Also dynamic loading calculation is a different long term stability issue compared to static load bearing. Go have a look at how the rail shims obviously work into railroad ties. A diesel engine is a rather dynamic load to accurately affix to a wooden structure.

[playdiesel]

You likely have done it but you can pull up the maximum allowable misalignmens on Lovejoy's web site, it aint much. Just be sure to read all of the material. There is no fine point of works vs does not work, all missaligments cost working life of the coupling starting at Zero. I have installed a few of them over the years and can say in my experiance with them in the field is if they need to act like a flex joint or universal joint then expect early failures of both the elements and drive hubs along with vibration. If the joint could almost use a solid shaft coupling then you will get reasonable life out of a Lovejoy as long as the torsional loads are reasonable..
I dont know enough about wood to say it will take "X" twobyes to make it ridged enough. I can say that when you have everything mounted down but with no coupling eliment grab the two shafts pulling up on one and pushing down on the other, then try sideways. If you can move things around perceptivelythen you are going to have coupling troubles, if not your in like flint.

[WStayton]

Ok, my engineer is showing here! <grin> I've been wrestling with the problem of figuring out what the loads are on my wood lamination topped with a piece of steel plate.

  The first question is what is the load.  First, I think it is safe to assume that the load from the engine far exceeds the load from the transmission or the generator, so that is what I will look at.  One way to look  at this, I think, is that the engine is only going to induce second order forces since it is primarily completely balanced with two pistons going up at the same time that there are two going down.  There is, however, some secondary forces since the pistons at both sides of the top of their stroke are moving slower than the pistons at the bottom of there stroke due to the fact that the crankshaft acts  like a lever and it is in a different configuration at the two positions.  How much is this second order force? Well, it's kind of hard to figure without knowing the weight of the pistons, the weight of the connecting rods, and the weight split (reciprocating vs rotary) of the connecting rods, and the length of the crankshaft throw. But, we can make an assumption about the LIMIT of how big this force is:  Since when the engine is sitting on something (think floor) and running, none of the four "feet" come off of the ground/base/floor, the force must be less than the weight of the engine.  Since this is a Mercedes engine, and they seem to believe that if some is good, more is better and a whole s$!t load is perfect, it weighs just a shade less than 500 lbs, so figure 125 lbs per "foot".  So the load is 125 pounds alternating on a 30.25 inch square of steel.  Since this qualfies as a "suddenly applied load" the effective loading is taken to be twice the actual loading (See any strength of materials text for an explanation of "suddenly applied load"), so we have an effective load of 125 pounds times load factor of two divided by the area of 30.25 square inches = 8.26 psi dynamic load alternating between this value and minus this value at twice the rpm as the engine rotates.  We have to add to this the static load of the weight of the engine which is 4.13 psi for a total max load of 12.39 psi in the down direction and 4.13 in the up direction.  If we assume that the steel block is rigid with respect (we will return to this in a minute) to the load placed on it, the questions is, is this load an elastic load for the underlying wood, or will it be plastically deformed.  Since the yield strength of wood is normally taken to be 6400 psi, we should be VERY safe with a safety factor of about 500 – normal machine design considers a safety factor of 3 to 4 to be sufficient for situations that are not dangerous to life and 10 to be sufficient for life threatening situations.  Back to the problem of the flexing of the steel plate.  Since the modulus of elasticity for mild steel (such as hot rolled) is normally about 30,000 psi, this means that the elastic deformation of the steel is 12.39 divided by 30,000 or about 4/10's of a thousandth of an inch (0.0004).  This makes the simplifying assumption that the stress is evenly spread out on the surface of the steel plate, which, of course is not true, but for a small load over a reasonable amount of area, the maximum stress and resultant strain is less than twice this amount, though the mathematics are too involved to go into here.  So, we will have a steel plate sitting on a wood support that is bending, at most 0.0008 inches, the questions is, is this within the elastic limit of the wood support.   The modulus of elasticity of soft wood is something like one thirtieth of steel, so it is around 1,000 psi so again the deformation would be 0.0008 divided by 1,000 or something like a thousandth of a thousandth of an inch which is well within the elastic limits of the wood, so the final answer is the load would not permanently deform the wood with a 0.250 inch mild steel plate of 30.25 sq in area. 

Note that this calculation only holds water if the weight of the engine exceeds the shaking forces so that the contact point of the engine does not leave the surface of the supporting structure.  If the engine is something like a Listeroid where the whole engine jumps up and down, clear of the surface supporting it, this whole method of analysis is invalid – you would have to look at something like impact analysis and for that the stress risers are MUCH more than the simple of factor of two for a suddenly applied load.

So, I think, that my 0.250 steel plate that is 5.50 inches by 5.50 inches, on top of a wood lamination is strong enough to prevent plastic deformation ("squashing") of the underlying wood.

I took a look at the problem of a railroad rail setting on a wood tie, and for that the analysis gets a little sticky since the wood is never firmly enough bedded in the dirt/gravel/substrate of the road bed to prevent the tie from moving when a locomotive axle weighing 40,000 pounds rolls over it.  Some industry measurements of a diesel-electric engine rolling along a track found that the axle of the train had a total movement of about 0.125 inches up and down as it rolled along over the track.  The generally held assumption is that half of this motion is caused by the tie moving in the substrate and the other half of it by the rail moving into the tie.   Since the axle has two wheels which rest on two different rails, that leaves a hold of 20,000 pounds being borne by one axle on one tie.   The rail is about three inches wide on its bottom (I think – I couldn't find any definitive dimensions so the three inches is a SWAG) and runs across an eight inch wide tie, the contact area is 3" by 8" = 24 sq inches.  This leaves a unit loading of 20,000 lbs divided 24 sq in or about 834 psi but this is again a suddenly applied load, so the load is effectively 1668 psi, 

I am not sure how to evaluate the strength of the wood in a rail road tie – since it is treated wood, and degrades with time, it must start out somewhere near the commonly held strength of construction timber of 6,400 psi, but as it ages it does get soft and railroad ties are replaced on a regular schedule for this reason. So, maybe by the time it is replaced it has degraded to 25% of this value, at least on or near the surface, which would make it 1600 psi, which accounts for the impressions that you can see by looking at a pile of old, replaced ties, where they have obviously been loaded plastic-ly and not elastic-ly .  Again, this isn't a very rigorous analysis, since the real load on a tie from the rail as a locomotive passes over it is not really documented – its mostly a matter of assumptions about the load bearing on ONE tie. A few pressure gauges would do wonders to firm up this analysis, but, so far as I can find, nobody has done that

That's the end of my strength of materials story (I would say analysis, but this is more nearly a cartoon than an analysis, so . . .), and I'm sticking to it! <smile>

Regardz,

Wayne Stayton

[Crofter]

It is not a question that the wood can take the load but it expands and contracts with temperature and humidity fluctuations. Where it is constrained against these forces (like being torqued down with steel bolts) permanent deformation gradually occurs after a number of cycles.

There is a reason for using steel shims to set motor and pump alignment instead of the much cheaper wooden wedges! You said it yourself; you are cheap.    I am quite frugal myself but I dont care to make fixes that dont stay fixed. Like I said, and I am sticking to that; I dont think wood makes good mounting medium for separate machines that require accurate and stable alignment.

[mbryner]

Hey, I liked reading all the engineering calcs you did.      I was a Biophysics major in college....   

Marcus

[WStayton]

Crofter:

  Ok, I'm gonna play devils advocate here.  How do you account for the hundreds, if not thousands, of diesel engines in marine service, that are/were mounted on something similar to what I am envisioning and haven't had many problems with alignment even though the shaft is usually fixed solidly to the drive shaft of the engine and complete constrained by the bearing at the through-hull fitting and stuffing box were it leaves the interior of the boat and is fitted with the propeller?

  Every wod boat ever built, that has a diesel engine, uses some for of timbers tied into the rest of the boats structure to hold the engine.  These timbers are frequently laminated, at least more recently, with large size suitable timber being harder to find. and they also frequently have metal plates on the top surface of the wood to spread out the engine load, though they are usualy "L"'s or channels of 0.125" thickness or little less and not 0.250" thick steel such as I am planning.

  The engine that I am going to use, came from such an installation  and used the four mountng feet that I am so enamored of with a prop shaft solidly mounted to the drivetrain and for the more than 2000 housr that it was installed I don't believe that it had a problem other than being less engine than the boat needed.

  And, actually, I got the idea for this type of mounting that I am planning from the many marine mountings that I have seen over the years.

  The mounting is completely constrained in a boat in that is built into the structure and I have only ever seen one mounting that was torn loose, but that was a 427  side-oiler with two four-barrel holleys on it driving a lightweight hydroplane.  When things "went a cropper" with the engine turning about 6,000 rpm, the mess was unbelievable - the engine came apart in not so big pieces during the course of the destruction and it appeared, after the fact that the engine swallowed a valve which caused an interferance fit between the piston and the valve and the head which caused the rod to break which left the engine so umbalanced that at that power setting it simply self destructed before anybody could pull the throttle.  The engine BTW was not a particularly durable setup - when we test ran the batch of 50 of them on a dynamometer, before delivering them to the customer, something approaching half of them came apart at WOT.  The crank bosses were thought to be inadequate for such high loads and this batch of engines was, more or less, a stop gap until a more robust batch of 4-bolt-main blocks could be cast and machined.  Needless to say, they were all returned and replaced, ultimately.  Of course WOT was something like 7,000 rpm which is really too many rpm for a large displacement v-8, but we had no way right at hand to limit the rpm so it was thought that the better side of valor was to test them at some rpm that they would never achieve in the boat, and if they survived that, they were "good-to-go". The theory worked fine, until one of them swallowed a valve!!! <grin>

  I agree that everything has to be water tight to minimize swelling and shrinking and I suppose that the environment should be of more or less constant humidity, not varrying between 90% relative humidity in the summer and 10% relative humidity in the winter.  Perhaps a marine installation has the edge here, since there is ALWAYS some water in the bilge and the engine room, of necessity, is less than toasty dry.  Maybe a I should put a pan of watter under the whole thing and contain it in a housing that will maintain the humidity?  I'm sort of kidding about the pan and housing but, I will completely seal every surface of the wood to minimize the shrinking and swelling caused by the uptake and drying out of the wood,  Oil based primer with a couple of coats of oil based enamel on top of that???

Regardz,

Wayne Stayton

[Crofter]

It is a good idea to seal the wood well. Somewhat of a V shape to the motor mount is generally valuable as not all the forces are straight up and down but if you can keep things from starting to move you should be OK. The generator will be a little tamer.  I would expect your proposed setup to need a bit of fine tuning to the alignment as everything settles in. If that happens to be the case it makes things a bit easier if you put twenty thou. or so of shims under each of the mounting points as part of the initial alignment.

Don't pull too many negative G's 

[WStayton]

Crofter;

  Don't think that negative G's are going to be too much of a consideration, since everything is going to be strapped/bolted down firmly to planet earth, and not doing zero G's apogees (sp?) in the "Vomit Comet"!  <grin>

  The alignment is going to be a time consuming process - I'm thinking that I will have two large format washers under each foot and the put much thinner, large sheet-metal shims, of varying thickness, as neccessary, between the washers.  Putting in the shims will be a PITA, though, since I think that they need to be pieces of sheet metal bigger than the large washers to ensure the load is spread evenly and the fact that they are sort-of square washers, will mean that I have to lift the engine off of the mounting bolts/studs everytime I change something!  I think that my engine hoist is going to get a work out - along with my torque wrench, since everything should, I think, be torqued down tight before any alignment measurements are taken.  Hmm, probably the most stressed piece of equipment in the whole chain is my arm doing the engine-stand-pumping and stud torqueing!!!  <grin>  Maybe I'd better start doing arm curls with weights? <smile>

  One thing I hadn't though about, though, is the alignment side-to-side.  I have just sort of sat here fat-dumb-and-happy, thinking that if the bolt-hole are drilled in alignment, everything else will be in alignment.  I reasonably confident that the holes in the engine mounting are in alignment, though the mounting was made by the French Company, Nanni, not by Mercedes, so maybe they are a little suspect.  The holes made in the feet on the generator-head, by some semi-skillled Chinese factory worker, using God-knows-what for a drill, with a drill bit last sharpened God-know-when, leaves them much more suspect and my design has no provisions for altering things if 1 or more of any of the mounting feet are slightly off except to measure them before hand and drill them off-set when I do put the holes in the wood and steel sandwich.  I was thinking to use a drill press to drill the holes, with the whole mounting rack blocked up to the right height so that I can get the drill press in to make the holes - but my holes are subject to error both in measurement and alignment of the drill to the measurements.  Anybody see a solution to this can-of-worms other than installing some adjustable apparatus to be part of the mechanism and exactly position the hole to fit? Sort of a micrometer poisitioned hole - which would degrade the quality of the stiffness of the hole with respect to the rest of the mechanism. Or am I again dreaming up a $500 solution to a $0.50 problem?

  Maybe just have the hole be 0.050 oversize then I could position the bolt in the hole, to get it where I want it and then clamp it down?  I am, of course, very shy of making the holes too big and then have a problem with holding the bolts securely enough to the mounting so that THEY won't move around. Ideally, the mounting bolts should be a sliding fit in the holes, not sure if that is achievable in this case, or not.  What does everybody think?

  Somebody, obviously, has been holding Listeroids down to such a substrate or, alternatively, to a steel one, what do you do for alignment in that situation?

Anyhow, thanx for your input!

Regardz,

Wane Stayton

[Crofter]

It is very hard to drill holes that precisely and really the bolt should not be expected to maintain the lateral position as it has very small bearing area in itself. You are then dependent on shear forces only between the contact plates and the wood so the torque has to be maintained.  Unless you can be assured they are bang on you might as well slot the holes and perhaps use some kind of auxiliary adjustable dogs on each side or a slide plate with set screws. Epoxy bonding of the plate to the wood? Plates recessed?

In setting pumps on concrete it is common to have space around the top 10" or so on each anchor bolt so they can be moved a fair bit for fine tuning then grouted after the millwrights did their final shim shining. You might conceivably do the same with epoxy between bolts and wood. Yes it is a pain to have to lift the units entirely off to add / remove shims; common to have them slotted but you give up some bearing area.

I used to shoe horses; now that is the ultimate in alignment dynamics and keeping a chunk of metal firmly attached to a flexing and yielding medium! It is a challenge. I dont subscribe to it but some people used to heat the shoe and hold it on for a few seconds to burn down any high spots to get better bedding. That stinks of sloppy fitting, pun intended, 

[mike90045]

  Somebody, obviously, has been holding Listeroids down to such a substrate or, alternatively, to a steel one, what do you do for alignment in that situation?

belt drive