gears analysis

sir i hided the pinion and gear separately and saw the stress generation the stress was concentrated at the centre i.e the middle portion

I  applied face meshing with everything fine i.e span centre and everything fine and smoothing high and high relevance at the contact region including both tooth and tooth space of pinion and gear i.e nearly  20 and 30 faces respectively  then i got these results

SIR I JUST WANT TO KNOW ARE THE RESULTS CORRECT SO THAT I CAN PUT THESE DATA IN MY REPORT AND PRESENTATION OR I HAVE TO APPLY SOME CHANGES

(off topic-sir regarding that faulty file i am working on that i will post with archieved file from now on very soon,i was trying to delete the messeage but whole conversations got deleted)

THE  face meshing i applied the details are here below

sir are these okay or if i apply more moment the will i get better stress deformation that will be visible well?

SIR ANY SUGGESTION?

 

«1

Comments

  • peteroznewmanpeteroznewman Member
    edited January 30

    As I said in the deleted discussions, you must do a Mesh Refinement Study.

    Make a plot that looks like the one shown below for each gear, then you can be confident that you have a proper mesh.

  • 5802lucky5802lucky Member
    edited January 30

    thank you sir but how to find that max stress and element size graph in ansys .What is the procedure? where is that option?Do i have to study the whole ASME V&V10.1 

  • peteroznewmanpeteroznewman Member
    edited January 30

    Maximum stress is on the legend.

    Element size is in the Sizing Mesh Control

    No, you don't need to read the ASME V&V10.1 guidance.

  • 5802lucky5802lucky Member
    edited January 30

    thank you  sir i will try with this i hope in this i will get nearly correct results

    Now sir i have been told to get the response at varying load under specific frequencies .I have done modal analysis of helical,herringbone and crossed helical gear pinion assemblies.I have kept the number of modes as 10 as told and have gotten different natural frequency wrt modes in total deformations.

    Sir now what to do in harmonic analysis,What should be my boundary conditions and load criteria.I have not been told in details about this.Should i just apply the same boundary conditions and moment as applied in static and transient or i have to apply by rotational velocity? I have watched some youtube videoes regarding this ,they soonafter modal analysis jump to harmonic analysis with copying the same model as modal analysis and then applied same boundary conditions and load  as static structural .Then what is the need of copying the total deformations of modal analysis to harmonic analysis? What exactly is the reasons of it?

    Then to perform this should i have to do it in explicit also? or harmonic is sufficient in my context

  • 5802lucky5802lucky Member
    edited January 31

    sir please give me ideas regarding the boundary conditions and load application of above mentioned

  • peteroznewmanpeteroznewman Member
    edited January 31

    What is on the other end of the shaft that has the gear?  How long is the gear shaft?  What is the gear shaft diameter? Where are there bearings?

    What is on the other end of the shaft that has the pinion? How long is the pinion shaft? What is the pinion shaft diameter? Where are there bearings?

    A Modal analysis is done on the system that includes the shafts, bearings and the gears.

  • 5802lucky5802lucky Member
    edited February 1

    but sir i have not been told to design shaft and bearing ,only to get the response under varying load under specific frequencies

    i have kept the shaft diameter as 1/4 th of pitch circle diameter as advised. Can it be done by only applying shaft not bearing, because i have not given any clue about bearings

    FOR MODAL ANALYSIS IS IT NECESSARY TO HAVE A GEARBOX? DO I HAVE TO MAKE A WHOLE GEARBOX FOR IT?

  • peteroznewmanpeteroznewman Member
    edited February 1

    Let's assume there is a bearing on the shaft close to the gear and the shaft is 200 mm long. Assume the same on the other side of the mesh for the pinion. The first natural frequency of that system is probably the shaft twisting and the gear and pinion rotating against each other with the teeth in mesh. Now make the shafts 300 mm long, the first natural frequency goes lower, or reduce the shaft length to 100 mm and the twisting mode natural frequency goes higher, but maybe that is no longer the first natural frequency. Do you see that you need to look at the system and not just the gear and pinion?

  • 5802lucky5802lucky Member
    edited February 1

    so sir is modal analysis possible without bearing? only taking shafts with variation in length?

     

  • 5802lucky5802lucky Member
    edited February 1

    And sir the rotational velocity option in load is applied on full body only.Will that give correct result in context of gear drive if remote displacement is applied in right dof for static analaysis?

  • peteroznewmanpeteroznewman Member
    edited February 1

    No, without the bearings the gear and pinion will flop about all over the place. The bearings are a required element of the system.  You don't need to model the bearing itself, just make a separate face on each shaft the width of the bearing, then use a Remote Displacement on that face. That way you can hold lateral displacements to zero, leaving all other directions and rotations free.

    What is the rotational velocity?  It is probably small enough to have a negligible effect. We are not talking about gas turbine rotational speeds here are we?

  • 5802lucky5802lucky Member
    edited February 2

    no sir i applied same torque i.e  in crossed helical and parallel helical gears, so the stress was higher in crossed helical ,it was acceptable because in crossed helical, there was point contact or line contact and due to the angle ,the faces in contact was 4 only ,where as in parallel shaft axis helical gear there was area contact and also the faces in contact was 5 (Both parallel and crossed helical gear-pinion assembly had same dimension only alignment was different)

    BUT  when i applied  a moment of 6000 rpm i.e 628.32 rad/sec in pinion of both parallel helical gear and crossed helical gear ,this time the result was opposite.This time with the same rotational velocity applied to the pinion body, the result reversed i.e the stress i.e max equivalent stress was higher in parallel helical gear instead of crossed helical gear

    SO  is in my context , rotational velocity applying to whole pinion body is correct procedure to compare stress?

  • peteroznewmanpeteroznewman Member
    edited February 2

    Please show your mesh refinement study plot and a close up image of the elements on the teeth for the smallest element size on the plot. Did you figure out how to plot the stress on just one body (without making the other body hidden)?

  • 5802lucky5802lucky Member
    edited February 2

    sir whenever i activate advanced sizing option on in mesh to kkep everything fine then my computer hangs, so i take coarse but i perform refinement with factor 3 on  all the contact tooth and some tooth around with face meshing.Sir i think the mesh refinement study plot is to be plotted by excel manually , you showd me where is the size of element but did not show me if there is any option in ansys to show the plot. So i ignored the plot , i think it will be done in excel i  will perform and show it to you.

  • peteroznewmanpeteroznewman Member
    edited February 2

    Don't use advanced sizing option. Use uniform sizing option with a coarse mesh size and a face sizing mesh control as I showed above.  Take the number used in the face sizing mesh control and the maximum from the stress plot and type those into Excel to make the plot.

  • 5802lucky5802lucky Member
    edited February 3

    sir there is no advanced option in face sizing details i have taken 5 mm is it okay. i took everything coarse and i also took refining 1 in the contact area along with this element size. 5 mm and face meshing in cont area also

    this is my conditions at cont area keeping other coarse

     

  • 5802lucky5802lucky Member
    edited February 3

    here i have taken element size from 0.5 to 5 and similarly stresses from maximum to minimum respectively i.e 10 datas in each axis

    In below i have taken from 0.025 to serially 10 value i.e 0.225 in x axis to stress from higher to lower

  • 5802lucky5802lucky Member
    edited February 3

    sir when i was doing that refinement with 3 factor it was taking huge amount of time and stress was coming higher, but after taking element size from  default to 5 mm and reducing refinement to 1 , it was solved within very few minute.So as only refinement with factor 3 alone takes huge time to solve can i assume that this gives the best nearabout result? In refinement 3 alone the stress was almost double than i got in refinement 1 anf face sizing of 5 mm .In refinement 3 alone the stress was 6e7 almost, now in refinement and element size 5mm stress max is 3.6e7 and iin only refinement 1 the stress was almost 2e7

    And sir another question-If i will slice both gear and pinion ,is it necessary to slice in same angle ? If i will slice gear and pinion randomly at different angle it will save time ,but will it give wrong results?

  • peteroznewmanpeteroznewman Member
    edited February 3

    Yes, these mesh studies take a long time.

    Don't use Refinement.

    Plot the individual points of stress for each Face Element Size.

    You can have a narrow angle for the gear and a wide angle for the pinion. Try to keep the circumferential length of each about equal.

  • 5802lucky5802lucky Member
    edited February 4

    Sir was my previous plot wrong?i made couple of plots as i was not sure. there are ten values of stress from minimum to maximum and the element size is 5mm i.e 0.005 m so in x axis what should be my range at what intervals .I  kept all the values from maximum to minimum in y axis serially fro bottom to top.Is it the right procedure?

    sir its okay if refinement with factor 3 tkes huge time , but will it give closely perfect results? Then its okay

Sign In or Register to comment.