linear or nonlinear dynamic analysis?

[Alessandro.ale]

Hello, everyone. just subscribe to the forum. I press that I have little experience with the software sw.
I am designing an optical access (silver + holder of the alloy steel window) to the combustion chamber of a motor. the system is then subjected, for each cycle, to a variable pressure load that lasts a fraction of a second and that has a peak of 50 bar (very lower than the material yielding loads). this load I have inserted it as a function pressure of time, with a lot of time curve and should be applied only once (not repeated, i.e. more cycles over time).

now, which of the linear dynamic and nonlinear dynamic simulations (which in my opinion are the most reliable for this case) reflect more in your opinion the reality?

I started, by curiosity, more simulations:
1) from linear static results were obtained positive results (i.e. optical access would hold the load);

2) dynamic linears: I have turned so many and with mesh more and more accurate and they have given me similar results, different obviously from static simulation (von mises much lower) but still positive.

3) nonlinear dynamics: high simulation time, in some cases they predicted von mises beyond the yield threshold for quartz, with in general values of von mises of 1-2 orders of magnitude higher than those of linear dynamic tests.

the problem is that the only thing that is non-linear is the load, the rest is all linear (material, geometry etc.)

Sorry, maybe I asked more than I could ask. Thank you in advance.

Say hi.

 

[PierArg]

hi alessandro and welcome to the forum,
I invite you to make a brief presentation!

 

[Alessandro.ale]

hi alessandro and welcome to the forum,
I invite you to make a brief presentation!

Hello. but me or my question?

 

[Stan9411]

of you. in the profile it says that you are student but clarify what you are doing helps to calibrate the answers.

Anyway, I'd like to think about it. In my opinion, the fact that you have a non-linear function load of time does not imply the need for a non-linear dynamic analysis. that type of solutors is necessary when measuring in analysis that present complex boundary conditions, force fields that are functions of dynamic quantities, need to evaluate any plasticizations of the material.
the typical example of all this is the simulation of crach tests: the load is non-linear function of time (before the impact there is, then ideally there is a step), the speed of the vehicle, the rigidities of the objects in play and many other parameters; you want deliberately to go beyond the ripple limit of the material, for this it is necessary to evaluate the non-linear behavior of the system.

now you have a panel, subjected to a known load history and, moreover, cyclic.
from my point of view:
- linear static analysis with maximum load
- static analysis to fatigue with maximum and average load to infinite life and criterion that seems more appropriate (the classic is gough-pollard)

Let me know what you think.

 

[Alessandro.ale]

of you. in the profile it says that you are student but clarify what you are doing helps to calibrate the answers.

Anyway, I'd like to think about it. In my opinion, the fact that you have a non-linear function load of time does not imply the need for a non-linear dynamic analysis. that type of solutors is necessary when measuring in analysis that present complex boundary conditions, force fields that are functions of dynamic quantities, need to evaluate any plasticizations of the material.
the typical example of all this is the simulation of crach tests: the load is non-linear function of time (before the impact there is, then ideally there is a step), the speed of the vehicle, the rigidities of the objects in play and many other parameters; you want deliberately to go beyond the ripple limit of the material, for this it is necessary to evaluate the non-linear behavior of the system.

now you have a panel, subjected to a known load history and, moreover, cyclic.
from my point of view:
- linear static analysis with maximum load
- static analysis to fatigue with maximum and average load to infinite life and criterion that seems more appropriate (the classic is gough-pollard)

Let me know what you think.

hello stan. thanks to the answer, I am a student in mechanical engineering of roma. I am currently doing the master's thesis on improving lean combustion processes (with excess air) and is the first time I use solidworks.

the rectangular prismatic glass of quartz that gives on the combustion chamber + holder of steel below (which does not overlook the combustion chamber) is not fatigued. the window rests on its steel holder and together they are in contact with the metal walls of the rectangular hole in which it is inserted. the combustion cycle is executed but once every 2-3 minutes, there is therefore a break during which the room is cleaned and the system is not subjected to loads.

the max temperature reached on the quartz window is on the 100°c, temperature that I don't think greatly affects its yielding load and stiffness, although on the internet I still haven't found much detail about it. So there's a thermal load, but I don't think it's so influential.

ok for nonlinear dynamic analysis, it was basically to try and see what happened

On the other hand, I thought I would use a dynamic linear analysis because I thought it was the one closest to the reality of the phenomenon. linear because the materials have linear behavior and then dynamic because so I could put a load function of time, which with static analysis I could not do. I have come to mind to put a dynamic load as the speed of application of the load influences the response of the material, which, however, solidworks I do not believe to take into account now that I think about it, i.e. I don't know what all its settings and parameters are to be put in the data of the material I suppose, but the fact is that where I shit can take such data? xd don't think they're available.

I also conducted linear static analysis with maximum load (even higher than maximum, for safety reasons) and the structure would hold.

I don't know how to move and how to rely on simulations if I make so many simplifications.