Non matching outputs - Dynamic Simulation

[asnair27]

This is a simple screw jack mechanism designed to study the Inventor dynamic simulation. There is a screw with a pitch of 2mm and is connected to a motor having 100 rpm speed which results in the linear motion of the top plate.

For simulation 100rpm (600 mm/sec) is given to the cylindrical joint:12 and 2mm screw pitch on Screw joint:13. Additionally a load of 2000N given at point P and gravity towards downward direction is also given. A calculation sheet is attached for reference.

The inputs given are as follows:

Link Length, L (mm)	200.00
Motor Speed, N (rpm)	100.00
Initial Alpha, α₀ (deg)	47.23
Screw Pitch = Lead (mm/rev)	2.00
Load on P (N)	2000
Direction of motion	Down

But the results were not obtained as per the calculation.

The initial position of the top platform is at 293.65mm.  

As per calculation, the height will get reduced to 119.9 mm after 16.5 seconds of operation. But in the simulation graph, its moved to 113.2 mm in just 0.035 seconds. Its not even driving for 1 sec.

What may be the cause of the error? How to rectify the problem?

[admaiora · Accepted solution]

Hi @asnair27 ,

the analysis is not giving the expected results because the kinematics are not set up correctly in terms of joints and other settings.
The analysis duration I can see (0.035 s) is extremely short, and the number of interpolation steps is also very low (6). There are serious redundancies in the model. Also, for this type of study, I do not really understand the need for 3D contacts, since they make the simulation more complicated and computationally heavier.
Last but not least, the mechanism is kinematically impossible as currently defined, meaning that the screw cannot work correctly in this configuration.
Complex mechanisms, especially in an initial study, should be tested progressively. To understand why the model does not work, simply try suppressing all 3D contacts, gravity, and the applied force. The relative translation on the non-threaded side must be removed. Otherwise, how could the mechanism work? Where would the axial force be reacted?

I cannot really correct your current model because it is too cluttered and contains too many joints. Reduce the model to the minimum number of joints actually required.
I am attaching a simplified model in which the results should match the expected values. I am using Inventor 2027.

Ciao!

[89198826955]

As I understand it,

you don't need the rotational speed;

you need to determine the force applied to the screw to lift 2,000 kg.

https://www.youtube.com/watch?v=zVvxvrsdYvw

[asnair27]

Thanks mate for the video. I think the translation motion was the issue. By locking the translation motion it worked perfectly as per calculation. 

I kept the 3D contacts, gravity and load still it works fine.

Some clarification for your kind information:

1. The analysis duration was kept very low because , the even if i keep 20s duration it was going to the end within 0.04 seconds.

2. The 3D contact was given to avoid interference between the parts.

One more doubt, if i give 10sec duration and 10 steps , it will gives values for each 1 sec intervals only, right? (Like 1s, 2s, 3s, ______ 10s)

But see the screenshot, the time interval is too low.  

[admaiora]

Yes, you can use multiples. 10 is a bit too low; with such a coarse time resolution, you might miss certain dynamic phenomena

[asnair27]

I think my question may not have been clear.

If I set the simulation duration to 10 seconds and select 10 steps, I would expect the results to be provided at 1 s, 2 s, 3 s, … up to 10 s, correct?

However, when I run the simulation, the time points are shown as 0, 0.00003, 0.00008, 0.00019, 0.00033, 0.00043, …, which does not appear to correspond to 10 equal time steps over 10 seconds.

Could you please clarify why the simulation is generating these time points instead of 1 s, 2 s, 3 s, … 10 s?

[admaiora]

Hi @asnair27 

This was caused by the contacts. They are useful, but you need to introduce them step by step.

You should adjust this setting.

Precise Events identifies the exact time when a specific event occurs during the simulation, such as a contact or impact, even if it happens between two output time steps.

[asnair27]

Thanks mate, that one helped alot.