Can you upload the model itself? It might be that the bolt connector is working properly and that the deformed shape "looks" wrong because of how much you've exaggerated the plot (although it's slightly also possible that the bolt connector isn't working). I'd also recommend that you check the bolt's reaction forces to see if it's carrying any load.
You probably already know this, but just to be sure, the second set of results with the radial constraints in no way represent real system.
Hi Shaun & Jared,
Thank you for your reply. I know that its not right to use radial (or circumferential) constrain (When I checked bolt reaction, due to that radial constrain, shear component on screws reduced). I just wanted to compare deformation & bolt reaction forces. When I used radial (or circumferential) constrain along with bolt connector, the maximum deformation was at two places (either side of center bolt) compared to just at center bolt, when I used just bolt connector.
So, I just wanted to make sure, bolt connectors working right.
Also, I have another question. As this is external pressure case, does bolt pre-load help any way? Generally we "hand-tight" in these cases, as external pressure help to keep things tight.
Bolt-check plot shows (I applied 80 in-lb pre-load for these 1/4-20 screws) that screws fail, but when don't use pre-load (just 0.1 in-lb so that I don't get that warning), bolt-check plot shows slightly better FOS for these screws. In either case, I am surprised that 1/4-20 screws are failing.
I thought having a tight lip (0.005 clearance) for the lid will help by supporting walls there by reducing shear on screws. But that didn't help the screws (.005 calculated FOS without lip Vs 0.007 with lip).
So I am not sure if my model working right.
I attached models (with and without lip).Could you take a look at these and advise me.
"As this is external pressure case, does bolt pre-load help any way? Generally we "hand-tight" in these cases, as external pressure help to keep things tight."
It depends. There are two factors in the real-life system that come into play:
- A change in the structure's stiffness due to the pre-stress from the bolt pre-load. This pre-stress may or may not have a meaning impact, depending upon the geometry and loading conditions.
- The higher the pre-load, the more frictional force there is to initially carry shear loads.
When we consider these two factors from the FEA side, then the question becomes whether or not the analysis is set up properly to capture these effects. I'm not sure if SW applies the bolt pre-loads first and then calculates a new stiffness matrix; if not, then the first item isn't taken into account. For the second item, this is only captured if your defined a friction coefficient and a "No Penetration" contact between the lid and the box and if (I believe) the pre-load is calculated prior to the other applied loads. Be aware that this will result in a non-linear analysis, which take longer to run.
"Bolt-check plot shows (I applied 80 in-lb pre-load for these 1/4-20 screws) that screws fail, but when don't use pre-load (just 0.1 in-lb so that I don't get that warning), bolt-check plot shows slightly better FOS for these screws. In either case, I am surprised that 1/4-20 screws are failing."
You need to keep in mind that these "bolts" are idealized representations. True bolted joints are non-linear and rather complex, so to reduce computational costs, we idealize the bolts as beam elements with some spider links between the beam and the neighboring solid elements. As I said earlier, a bolt per-load will only help reduce the shear load in the bolt if you account for the friction between the two faying surfaces. Also keep in mind that you need to make sure the calculated values have converged (I noticed you're using a draft quality mesh, which is good for de-bugging a model but bad for accurate results).
"I thought having a tight lip (0.005 clearance) for the lid will help by supporting walls there by reducing shear on screws. But that didn't help the screws (.005 calculated FOS without lip Vs 0.007 with lip)."
Tiny details like this are hard to account for unless the model has been set up properly. This includes capturing non-linear effects such as friction, as well as geometry changes. A 0.002" change may not have an effect if you're mesh is too coarse to capture this change.
"So I am not sure if my model working right."
I think that your model is working right insofar-as the bolts are carrying load (as opposed to not). I ran your model and checked the connection loads and all bolts are definitely carrying load. I'd spend more time on trying to get the faying surfaces to carry a reasonable amount of shear load via friction however.
Also, applying the pressure load on the same surface as the fixed geometry meaningless; I'd change the pressure load to exclude the bottom surface. You should also consider whether a fixed constraint give s a realistic representation of how the real system is held in place. Is this box under hydro-static pressure (i.e. under water)?
Thank you so much for your suggestions.Yes, it is under hydro-static pressure. I should have given fixed fixture to bottom 4 mounting holes. I started with a quick study (thought of updating mesh quality, fixed restrain, pre-loads ..etc later) and updated few settings during iterations. Will run the study with fine tuned settings and see how it goes.
bolt plot looks good to me, i don't think i understand why you think it looks wrong.
bolt and the radial displacement aren't the same at all. remember that a restraint is relative to the environment. not relative to each other. you'd need a pin connector to be closer to a bolt.
How to replicate approximate screw connection(without bolt connector) with boundary conditions with pre-load and without pre-load?
Did you use a coefficient of friction in the no penetration contact between the lid and box? The clamp force from the bolt pre-load wont do any good unless there is friction between the parts being bolted.
I didn't set friction in my original Study. Adding 0.5 friction between those two "No Penetration" surfaces did help "calculated FOS" X 10 times (0.006 FOS without friction Vs 0.07 with Friction). By using 0.25 thick lip with 0.005 clearance with box walls, helped FOS a little bit more to 0.2.
why do you want to replicate the bolt? the closest thing would be a pin with some kind of spring force to pull them together or pin with equal opposite force