topic Re: Applying Force and acceleration in Simulation Mechanical Forums (Read-Only)

Applying Force and acceleration

Anonymous — Fri, 15 Jul 2016 16:17:59 GMT

Hello All,

i´m trying to apply a force to a beam which can slide above a curve in horizontally direction and makes for that a displacement in vertically direction along a curve.

Since i apply the force on the curve and fix the beam at the other side, the curve "breaks through" the beam. If i apply now the force onto the previously fixed side of the beam, and fix the curve now, the analysis stuck at the timestep zero.

The background is, that the sliding of the beam across the curve happens in the reality with a certain value of acclereation. I have to determine, if my concept of beam and curve is accurate for this accerealtion.

In other words: i have to see, if my beam takes off the curve while sliding above it.

Since i can´t just apply a value of acceleration at the beam (..or havent find out yet) i have to apply a force with the realationship F=m*a with "a" as the acceleration- value and "m" as mass of the moving part.

Following picture shows the system, with a fixed curve and a beam which can just sliding in x-direction:

My steps for this dynamic investigation are:

1) Get the displacement of the free end of the beam in vertically direction:

2) With the second derive of displacemenet i get the acceleration of the curve:

3) With the curve of acceleration, i have a load-curve, where i can input my acceleration value in x-direction on the beam end. If the new displacement in vertically direction of the free end of the beam is bigger than the curve in step 1), i know that the beam takes off.

My questions are as follows:

- why breaks my part into the other part by apply a force?

- can i just apply a acceleration-value or do i have to do this detour with F=m*a

- are my investigations rigth respectively can MS handle my prolbem?

At the end i attached a .fem data.

In the first scenario i did just a displacement to geht the curve. In the second displacement, i isolated the beam an gave the beam at the free and a verticaly displacement like the curve is given and a horizontally force like the acceleration-curve is given. Unfortunatelly it doesnt work..

The third scenario is a multibody scenario like i explained at the beginning..this doesnt work either.

Hope you guys can help me?

Thank you very much!

Re: Applying Force and acceleration

John_Holtz — Fri, 15 Jul 2016 18:05:34 GMT

Hi Tomatello,

Here is an answer to the question "how to apply an acceleration". You can calculate the displacement at any instant in time from a known acceleration (either by a numerical integration or a formula), so you can simulate an acceleration by using a prescribe displacements. Please see the documentation for the prescribe displacement; it discusses entering a velocity or an acceleration. For an acceleration, the documentation discusses what you need to be aware of in case the analysis reduces the time step.

Hope this helps. I will try to look at the models to see what is occurring with the results.

Re: Applying Force and acceleration

John_Holtz — Fri, 15 Jul 2016 21:27:30 GMT

Hi Tomatello,

I think using the prescribe displacement to move the beam is the way to do the analysis. See my previous post.

In design scenario 2 and especially in design scenario 6, I suggest that you replace the pin (part 4) with a joint ("Mesh > CAD Additions > Joint"). If you select the surfaces of the hole on the wheel (part 3) and beam (part 2), the software will create the lines needed to simulate the pin. This is a much more efficient way to allow the wheel to rotate than to try to simulate the pin and surface-to-surface contact.

The wheel will not rotate the way you have the model setup. To understand why, think of a case with a square pin passing through square holes with 0 clearance. The wheel will not be able to rotate. In your model, you have a round pin and round holes, but I believe the clearance is zero or close to zero. Because the pin and hole are approximated with polygons, they will not be able to rotate (just like my square pin and hole example). In fact, you can calculate how much clearance is required depending on how many sides are used in the mesh of the pin and hole. (See the page "Advanced Tab" in the documentation. Figures 3 and 4 show the problem with the approximation of a round shaft in a round hole.)

Another option you might want to consider is to lower the curved wedge (Part 1) a little bit in the Y direction so that the wheel tries to move above the wedge instead of moving under it. It looks to me that the model is setup so that the wedge impacts the wheel directly through the center of the hole, so the beam is just as likely to deflect down as it is to deflect up. Also, this "knife edge" contact is probably hard for the solver to converge; moving the wedge down 0.2 mm will help with this initial impact.

Good luck.

Re: Applying Force and acceleration

Anonymous — Tue, 26 Jul 2016 19:48:07 GMT

Sorry for the late response. The hint with the joint was a good one!

And thank you for the other hint with the acceleration..

Let me summarize it:

1) I have my curve, which passes a distance of x = 7.5 mm in a time of t = 1s

The acceleration is for that a = 2*x/t^2 = 15 mm/^s.

2) The acceleration i give the curve on the plane side in the direction to the fixed end of the beam is: x = 0,5 * a * t^2 = 0,5 * 7.5mm * t^2. Since t is the timestep, it is my control variable and i get the following plot:

My question here is: do i have doubling the plot respectively the multipicator to get in the first timestep: 2 mm* 7,44 ~ 15 mm/t^2 ?

3) now i can see (maybe i have to enable the extendet timesteps) if my wheel loses the contact to the curve.

Another interesting point is, if i can set displacements, which acting like the "bearing load" option. For an example: a bearing load at the end of ma beam would imply a displacement wand the beam makes a circular arc. If i make a prediscribe displacement in the y-direction, the bearing moves only in the the y-direction and at the higher it gets, the more the beam get streched.. i hope you understand what i mean?

Re: Applying Force and acceleration

Anonymous — Tue, 26 Jul 2016 21:02:46 GMT

I think i can answer my first question by myself. If i run the simulation with the doubled value in the checkbox of "Größe", my contur runs 15 mm. Since the distance of the curve + the point at the joint is 7.5 mm i have to leave the value at the checkbox at 1 mm. Otherweise the wheel would run above the curve and drops at the plane side of the curve to the ground (x-axis) and would run another 7.5 mm .. but this would mean, that the acceleration is always the same to get a displacement of 7.5 mm??

Re: Applying Force and acceleration

John_Holtz — Wed, 27 Jul 2016 12:21:02 GMT

HI Tomatello,

I do not understand the motions that are involved.

In item 1), you calculated that the acceleration of part 1 is 15 mm/s^2. So part 1 (see figure below) starts at 0 velocity and accelerates at a constant rate of 15 mm/s^2. Correct?

In item 2), you wrote x = 0.5*7.5 mm/s^2*t^2. The 7.5 mm/s^2 should be 15 mm/s^2, correct? Perhaps you meant x=7.5 mm^2*t^2.

The next thing that I do not understand is the plot of the load curve. That is not a graph of x=7.5 mm/s^2*t^2. It looks more like a curve for a body that is initially moving at some velocity and decelerating to a stop. If this is the case, then the equation in item 2 is wrong. (The full equation would be x = x0 + v0*t+0.5*a*t^2 where x0=0 and v0=the initial velocity.) If part 1 has an initial velocity, you might want to apply that as a load just to make the setup "look" more complete. It will not have much effect in this case because part 1 is so compact and the prescribe displacement will create the initial velocity.

Re: Applying Force and acceleration

Anonymous — Wed, 27 Jul 2016 20:53:49 GMT

Hi John,

i appreciate your effort 🙂

Item 1): The curve starts at t0=0 with v0=0. The picture you have showed describes the beginning task. I just know, that the curve arrives the end-distance of x=7.5 mm i a time of t=1s. At the end of the task, the wheel is on top of the curve, where the axialy axis of the wheel-beam (center of the wheel) is in one line with the plane side of curve on the vertically.

So for that, i assume to calculate the acceleration with a=(2*x)/t^2 -> 15 mm/s^2. So yes, this is a constant acceleration i guess..

ITem 2) Yes, you are right! The equation is: x(t) = 0.5 * 15 mm/s^2 * 1s^2. And yes you are right with the plot: in this case i muddled up the x-accelereation with the y-acceleration (see my pictures in my start-posting).

The real plot is shown below:

Re: Applying Force and acceleration

John_Holtz — Thu, 28 Jul 2016 12:06:02 GMT

Hi,

Great. Looks like you have everything setup properly.

To answer your question at the end of item 2 (in your original email), the magnitude of the prescribe displacement on part 1 (the circular wedge) should be 1, and the load curve then defines the total displacement of the wedge versus time.

In your original question 3, you were asked about the prescribe displacement (PD) at the wheel-end of the beam to simulate the wheel moving over the wedge. To summarize the question, does a PD in the Y direction cause the beam to get longer because it restrains motion in the X and Z? You have the option to control that!

If the PD is set to move in the Y direction (with either a + or - magnitude), then the PDs do not restrict motion in the X and Z. So as the wheel is pushed higher and higher in the Y, the beam is going to curve in the X direction so that the length of the beam remains (almost) the same.
If the PD is set to move in a custom direction, such as [0,1,0] or [0,-1,0], then you are telling the PD to move the nodes 0 in the X, +/-1 in the Y, and 0 in the Z. In this case, the beam has to stretch as the wheel is lifted in the Y direction.

Does this make sense?

Re: Applying Force and acceleration

Anonymous — Sun, 31 Jul 2016 01:58:58 GMT

Hi John,

yes this makes sense to me!

In the meantime, my task chaned a little bit..

The acceleration to the beam or the curve is now step by step with the following load-task:

At the end of this load curve, the task shall repeat the load-curve till the wheel is on top of the curve.

But i dont know how i can reapeat load-cuves??

I hope you can help me?

Tanks!

Re: Applying Force and acceleration

AstroJohnPE — Sun, 31 Jul 2016 02:25:20 GMT

I would export the load curve to a text file. Then use Notepad or Excel to create as many repeats as you want. Then import the text file into the load curve.