Analysis on log splitter built

[polemidis]

I would never guessed, but look what a HUGE difference the Finite Element Analysis results to these models.
I applied 35000Newtons to the cutting edges, with the cylinder base pin location as the constraint (the red round marks at the right of the pictures)


The one at the left shows a distortion to the Z axis of 0.4mm, while the one on the right 5.5mm! In Von Mises streses the left gets 158MPa of pressure, the right one 370MPa. With yield strength of steel of 250MPa that means that the right one will deform in that extreme case.

What is the difference? The left one has the has the cylinder mount 2in inside the I-Beam, while the right one has it flush to the end.

 

[sostahoe]

What about an extention of the beam on both ends? How much extension before the stress reduction is disapated before beam ends.

 

[polemidis]

What about an extention of the beam on both ends? How much extension before the stress reduction is disapated before beam ends.

I tested only one end. I think 2" in ok. For the other end I may add 1 inch, since there is grate mounted that reinforces the beam more.
That splitter is part of a firewood processor.

FYI guys the beam tested was W flange, 6.5sides with 8" web. The actual one that I will install is 8" flanges with almost 7.5"web and 1/2" thickness, skrutzed from craiglist. I thick I cannot go wrong with that I

 

[4570Man]

There's more forces involved in a splinter than that. That's failed to account for what happens when a piece of wood is cut crooked or a knot hits one side of the 4 way beams. I'd overbuild the beam by a good margin.

 

[s219]

I would never guessed, but look what a HUGE difference the Finite Element Analysis results to these models.
I applied 35000Newtons to the cutting edges, with the cylinder base pin location as the constraint (the red round marks at the right of the pictures)
View attachment 542116View attachment 542117

The one at the left shows a distortion to the Z axis of 0.4mm, while the one on the right 5.5mm! In Von Mises streses the left gets 158MPa of pressure, the right one 370MPa. With yield strength of steel of 250MPa that means that the right one will deform in that extreme case.

What is the difference? The left one has the has the cylinder mount 2in inside the I-Beam, while the right one has it flush to the end.

When I sanity check those two cases, it doesn't pass my eyeball tests. Either something else changed that you're unaware of, or the boundary conditions are wrong, or even the color scales aren't consistent between the two. I would not expect that much difference just by moving the cylinder mount like that.

 

[polemidis]

There's more forces involved in a splinter than that. That's failed to account for what happens when a piece of wood is cut crooked or a knot hits one side of the 4 way beams. I'd overbuild the beam by a good margin.

Absolutely. The analysis is done to one step at a time I just posted it because I found it interested. Not as a complete analysis.

The beam I am using I think is "overbuilt" no? W8x40 for W-Beam is pretty descent imho. I may not even bother to brace the ends

 

[polemidis]

When I sanity check those two cases, it doesn't pass my eyeball tests. Either something else changed that you're unaware of, or the boundary conditions are wrong, or even the color scales aren't consistent between the two. I would not expect that much difference just by moving the cylinder mount like that.

I was surprised too, thats why I bothered to post it. Nothing else changed. The only think that I would suspect, is to the creation of the mesh?? Anyway, I would never know. I have updated the model to my real size beam (W8-40) that I bought, which is bigger than the one I had modeled above. I would run a test again to see what shows now.
Can you run an analysis? I can sent you the .step file, I was wondering if the results vary a lot

 

[LD1]

I cannot begin to count the number of times I have seen a home-built spitter rip the wedge or cylinder anchor off the beam because it was only "surface welded" to the flange of the beam.

Huskee store bought splitters, which are some of the cheaper (but still good) ones out there dont even do that.

Cut a pocket and insert the anchor block and wedge.....see what that does for your numbers

 

[polemidis]

I cannot begin to count the number of times I have seen a home-built spitter rip the wedge or cylinder anchor off the beam because it was only "surface welded" to the flange of the beam.

Huskee store bought splitters, which are some of the cheaper (but still good) ones out there dont even do that.

Cut a pocket and insert the anchor block and wedge.....see what that does for your numbers
View attachment 542171
View attachment 542172

HA! I got you! I have already done that! Its at the other side not visible from this view. Actually to say the truth and nothing but the truth, I cut the pocket mainly because of the need to adjust the 4way with hydraulics below the beam, and then realized that it helps A LOT. But I never thought of doing that to the cylinder mount!!!!! Even though its half the pressure (since it is 2 "ears") its cheap insurance. (And an excuse to put the plasma to work)!!!!

So this where I am right now. Wtf, I have so many designs now. But no money to built any of them!!!!:confused3: Even though I figured out the cutting system, I removed it for now, since the spring is almost here, and I have to start cutting. The the more risk-free way is to mount my spare chainsaw and do it manually. Anyways it saves me huuuge time and effort of making firewood.

 

[texasranger556]

Agreed. At a minimum your color scale changed. I can't believe that moving the cylinder mount left 2 inches on the beam would have any affect at all on the left edge of the 4 way, and yet it turned from blue to red. Something isn't adding up.

Edit: this was meant to be a reply to s219

 

[KB7722]

Box the beams on each end. Or maybe just an end plate would work. Basically the load is compressing the beam Web on each end beams aren't designed to be loaded such.

 

[LD1]

Agreed. At a minimum your color scale changed. I can't believe that moving the cylinder mount left 2 inches on the beam would have any affect at all on the left edge of the 4 way, and yet it turned from blue to red. Something isn't adding up.

Edit: this was meant to be a reply to s219

It's how FEA tools work.

You gotta look beyond the red to blue color switch. You gotta look at the raw data.

I like to look at the stress.....but he has chosen to display the "displacement" or deflection.

The fact that the 4-way end went from blue to orange means nothing. In fact.....the deflection at that point is exactly the same on both cases.

What happens is the FET tool automatically scales it for you. And you can see in an instant where the weakest link was.

In his first pic....with the anchor block at the end and lit up red.....it shows a displacement of -5.5mm.

In the second, where it shows the 4-way being orange.....it shows a displacement of -0.4 mm.

So even in the first one....that wedge was displacing -0.4mm.....but in the scale.....to encompass the -5.5mm on the anchor block......the wedge area shows up as blue.

So you gotta understand scale when using an FEA tool. Orange and red just show the weakest area of the whole structure.....but that doesn't mean that area is actually weak and/or gonna fail....because you have to look at raw data.

Kinda like if you weigh 200#.....and you step on a 200# scale......you just pegged it into the red. But if you step on a 2000# scale....the needle would hardly move and you would still be in the blue.....your raw weight (raw data) didn't change. Just the scale at which you were measuring with

 

[polemidis]

It's how FEA tools work.

You gotta look beyond the red to blue color switch. You gotta look at the raw data.

I like to look at the stress.....but he has chosen to display the "displacement" or deflection.

The fact that the 4-way end went from blue to orange means nothing. In fact.....the deflection at that point is exactly the same on both cases.

What happens is the FET tool automatically scales it for you. And you can see in an instant where the weakest link was.

In his first pic....with the anchor block at the end and lit up red.....it shows a displacement of -5.5mm.

In the second, where it shows the 4-way being orange.....it shows a displacement of -0.4 mm.

So even in the first one....that wedge was displacing -0.4mm.....but in the scale.....to encompass the -5.5mm on the anchor block......the wedge area shows up as blue.

So you gotta understand scale when using an FEA tool. Orange and red just show the weakest area of the whole structure.....but that doesn't mean that area is actually weak and/or gonna fail....because you have to look at raw data.

Kinda like if you weigh 200#.....and you step on a 200# scale......you just pegged it into the red. But if you step on a 2000# scale....the needle would hardly move and you would still be in the blue.....your raw weight (raw data) didn't change. Just the scale at which you were measuring with

You are correct. I put the stress numbers on my first post though. Not able to confirm if was something wrong with the mesh creation, this is how it looks now. Be aware that I have not knowledge nor experience on using FEA. I just learn from google here and there.
So if you think that I am using it wrong plz let me know. This current analysis has changed: a) The flange width is 8" instead of 6.5" , and I created pockets so the cylinder mount "ears" and at the 4-way knife can go in and be welded to the full length of the web, plus at the opposite flange.
One pic shoes the Von Mises stress, the other the Tresca Stress. I have no idea what these mean though, if anyone can help me PLZ DO! Resources to read, or even a plain explanation, anything that helps me.
The Z displacement is 2mm, which I believe is elastic deformation. Applied force is 35000 Newton

 

[LD1]

Basically my last post was in attempt to explain how the FEA tool works. As one was questioning why all of the sudden the wedge went from blue to red...(seeming thinking it got weaker)

It did not.

FEA tool is always gonna have green and blue all the way to red and yellow. It don't matter weather you have a 40" tall bridge beam and put a mere 10# load on it. Some point is always gonna be red. Because it automatically scales unless you manually set the scaling.

Hope that makes sense

 

[texasranger556]

Basically my last post was in attempt to explain how the FEA tool works. As one was questioning why all of the sudden the wedge went from blue to red...(seeming thinking it got weaker)

It did not.

FEA tool is always gonna have green and blue all the way to red and yellow. It don't matter weather you have a 40" tall bridge beam and put a mere 10# load on it. Some point is always gonna be red. Because it automatically scales unless you manually set the scaling.

Hope that makes sense

I guess my point was you likely can't compare the two pictures absolutely because the color scale is relative to each, as you confirmed in your explanation. Thanks for that.

Matlab does the same thing by default, but you can set color scales manually, change to logrithmic, etc depending what you're trying to show or compare.

 

[LD1]

I guess my point was you likely can't compare the two pictures absolutely because the color scale is relative to each, as you confirmed in your explanation. Thanks for that.

Matlab does the same thing by default, but you can set color scales manually, change to logrithmic, etc depending what you're trying to show or compare.

Correct. Two pics dont compare because scaling is different.

You said:

I can't believe that moving the cylinder mount left 2 inches on the beam would have any affect at all on the left edge of the 4 way, and yet it turned from blue to red.

I was simply trying to describe WHY it changed from blue to red. Because the reality is the hard data didnt change. And no the cylinder mount didnt have any effect on the 4-way. The change that was made forced the FEA tool to re-scale......thats why it changed from blue to red. But in BOTH pics, the forces and stress on the 4-way are exactly the same

 

[texasranger556]

Correct. Two pics dont compare because scaling is different.

You said:


I was simply trying to describe WHY it changed from blue to red. Because the reality is the hard data didnt change. And no the cylinder mount didnt have any effect on the 4-way. The change that was made forced the FEA tool to re-scale......thats why it changed from blue to red. But in BOTH pics, the forces and stress on the 4-way are exactly the same

Understood. I've never used FEA software to know they auto scale every time you plot, which you've cleared up for us. Thanks.

 

[s219]

You are correct. I put the stress numbers on my first post though. Not able to confirm if was something wrong with the mesh creation, this is how it looks now.

Yeah, the previous problem was much more than behavior of the color scale -- something was definitely wrong based on the numbers you quoted. But when you are comparing multiple cases, be sure to set the color scale manually so that it's consistent. That way you can make reasonable conclusions about which design is better. This is especially important when dealing with stress concentrations.

 

[DaveOmak]

Thank you.... All of this is WAAAAY over my head... BUT... I do understand the concept of inserting different parts of the splitter into the beam... It all makes sense... Over build everywhere... I planned on boxing in the beam and using it as an oil reservoir also....