RedJaron :
I'm very familiar with the difference. But nothing I said was incorrect. Compression and torque are very different. I simplified it for anyone that doesn't have an engineering background or degree ( which doesn't include me. ) You said "why is 2 pounds of force an issue when the recommend clamping force for TIMs is 55- 70 pounds on the heat sink / water block posts ?". And my answer is still the same. Because the forces from moving the machine around with a huge cooler tower drastically changes the forces from when it's at rest, especially if the mboard is vertical.
Now yes, you get a little bit of shearing strain at the mounting pins due to the curvature of the board. But again, that's a static load, and unless you're sliding the box back and forth along a desk, the forces at each mounting post will be roughly even. That's not the case when the board is mounted vertically, which you acknowledged. That's all I was getting at. And yes, I've transported my machine a lot as well without cracking the board. No one's saying you can't. I'm just saying it's a consideration. I can't guarantee everyone handles their machine as carefully as I do mine. What's rough handling for some is walking on eggshells to another, so I'm not going to assume that everyone knows how to transport a tower cooler and treat it as though it's not a risk.
I'm not arguing the transportation issue which was not part of the original premise. The question is the relative impact on the MoBo of air versus CLC coolers on the desk. The statement used as the premise is "the 2 pound load hanging off the MoBo" being a concern. Transportation is another issue entirely. Again lets look at the statement you quoted.
""why is 2 pounds of force an issue when the recommend clamping force for TIMs is 55- 70 pounds on the heat sink / water block posts ?"
A dropped box doesn't fit under that scenario. A dropped box will created a impact loading far, far greater than 2 pounds and that's why it's not part of the premise of the post, but a separate discussion.
As for the rest, starting with "shearing strain at the mounting pins due to the curvature of the board...."
Again, no. Bending and shear are two very different things. Curvature results from the top of the board being in compression and the bottom being in tension. This is not "shear". In the above beam image, shear is a force that exists in the vertical plane only. The 70 pounds of clamping force is resisted by the upward force of the mounts. In a typical "I" beam, shear is resisted by the cross sectional area of the "I".
Shear is a sliding failure on a material along a plane that is parallel to the direction of the load. The shear stress is consistent from one end of the beam to the other, and failure may theoretically occur at any point along it's length as the stress is the same throughout.
In the same "I" beam, bending stress or moment is more complex; the top flange of the "I" is in compression, the bottom flange is the opposite as it is in tension. In other words, the
bending stress is a failure in a plane that is perpendicular to the direction of the load. Examples:
Cut a branch with pruning shears = Shear Failure
Break a branch over your knee = Bending (tension stress) failure
Another reason why these two are very different ...
Resistance to shear failure doesn't change with the length between supports and loads
Resistance to bending stress failure increases exponentially as the length increases
Were you to over apply the clamping force to the CPU tot he PCB's breaking point, the top of the board would be in compression which the material will handle quite well. The bottom of the board will be in tension and this is where the crack will occur and failure will always be at the mid point directly under the load in a simple beam situation. On a MoBo, the presence of a socket and sharp corners in the PCB opening where it is inserted is likely to produce stress concentrations at these points.
Having practiced the engineering profession for 40 years, owning an engineering consulting practice for 25 and having taught engineering at the college level, I'll just say that shear, torsion and bending are all different things. On your desk, the 70 pounds clamping force of the heat sink creates a strain on the motherboard orders of magnitude above what it experiences from a two pound **static** weight.
You introduced the transportation issue which is a whole 'nother discussion unrelated to the above premise. That discussion would include impact loading and momentum which do not exist on your desk..... when being transported, the weight is no longer static,
I agree that shipping or dropping a box is always going to be a scary situation and for many reasons besides the weight of the heat sink. How long was the water cooler sitting in the truck in freezing temperatures ? Is the 'anti-freeze" adequate for those temps ? My box has a half gallon of coolant in it.... that's over 4 pounds ... no concern there ? Are the connections still water tight after all that bouncing around ?. Of course proper precautions must be taken which mitigate these concerns
These guys ship very complex systems all over the place
https://www.youtube.com/user/SingularityComputers
Bracing the heat sink with guy wires and / or styrofoam works well with large heat sinks. But again this is a separate discussion which really shouldn't need to take place if you ship responsibly and take proper precautions.
So, sticking to the original premise, that being "air or CLC cooler on the desk" ....
Clamping force = significant load on MoBo
Static weight = insignificant load on MoBo ... and regardless of whether mounted horizontally or vertically.
I have had users bring in non-functioning boxes for diagnostics over the years. The modes of failure observed were as follows:
1. Thermalright Silver Arrow / Asus Sandy Bridge WS MoBo - The Heat sink was over tightened, resulting in what I believe was warping of the socket which caused the socket pins to lose contact with the CPU pins, simply loosening the hold down screws returned system to proper operation.
2. Hyper 212 / Asus 1156 board- One of washers on the clamping mechanism had a burr on side facing the MoBo which resulted in scoring of the MoBo surface when mounting thumbscrew was tightened down. Burr scratched MoBo surface apparently cutting a circuit trace in the board.
3. Hyper 212 / Asus 1155 Board - Same issue as Silver Arrow
4. Hyper 212 / Asrock 1156 board - Board was cracked (bending failure)
On a side note, situations 2 and 4 required new 1155 MoBos (1156 was out of production) and CPUs and both times Windows activated w/o issue.