## Timing Results, Kinematics Analysis

Post any timing results you get for the Kinematics Analysis in HW 10 here.

### Re: Timing Results, Kinematics Analysis

r-w = 2078.021 s (~35 min)

r-p = 1778.619 s (~30 min)

r-p = 1778.619 s (~30 min)

### Re: Timing Results, Kinematics Analysis

ME751Anne wrote:r-w = 2078.021 s (~35 min)

r-p = 1778.619 s (~30 min)

are you getting the same results in both cases? I hope you do.

And if you did, it means that you got it right. In which case, you should have answered your own post that i just answered two minutes ago. I see no flaw in my logic...

dan

### Re: Timing Results, Kinematics Analysis

For 10000 calculations (0.001s)

r-w = 416s

r-p = 481s

For 500 calculations (0.02s)

r-w = 24.5s

r-p = 28s

I didn't include rendering times, just calculations of position, velocity, and acceleration.

r-w = 416s

r-p = 481s

For 500 calculations (0.02s)

r-w = 24.5s

r-p = 28s

I didn't include rendering times, just calculations of position, velocity, and acceleration.

### Re: Timing Results, Kinematics Analysis

for step size 0.02 for 10 seconds:

r-w formulation 6.6s

r-p formulation 6.1s

only calculations of pos vel and acc of both points p and q

r-w formulation 6.6s

r-p formulation 6.1s

only calculations of pos vel and acc of both points p and q

### Re: Timing Results, Kinematics Analysis

I used h=0.01 for 10 seconds of simulation.

r-p: 205.72 sec

r-w: 205.64 sec

Looks like I am slow but consistent.

r-p: 205.72 sec

r-w: 205.64 sec

Looks like I am slow but consistent.

### Re: Timing Results, Kinematics Analysis

ME751Toby wrote:I used h=0.01 for 10 seconds of simulation.

Toby - i think you should have had h=0.02, that would have cut the CPU time in two...

dan

### Re: Timing Results, Kinematics Analysis

step size: 0.02

r-p formulation: 24.379 s

r-w formulation: 26.141 s

r-p formulation: 24.379 s

r-w formulation: 26.141 s

### Re: Timing Results, Kinematics Analysis

r-w =385s

r-p = 352s

r-p = 352s

### Re: Timing Results, Kinematics Analysis

r-w: 60sec

r-p: 65sec

r-p: 65sec

### Re: Timing Results, Kinematics Analysis

r-omega formulation: ~30 minutes

r-p formulation: ~34

These numbers should be taken with a gigantic grain of salt as I've just realized my laptop crunches matlab code at about ~1/10 the speed of a CAE computer. Go figure.

On a reasonable computer, I'd guess closer to 4 and 5 minutes, respectively.

@Anne: Glad you measured your time out in milliseconds.

r-p formulation: ~34

These numbers should be taken with a gigantic grain of salt as I've just realized my laptop crunches matlab code at about ~1/10 the speed of a CAE computer. Go figure.

On a reasonable computer, I'd guess closer to 4 and 5 minutes, respectively.

@Anne: Glad you measured your time out in milliseconds.

### Re: Timing Results, Kinematics Analysis

Rob Bradford wrote:r-omega formulation: ~30 minutes

r-p formulation: ~34

These numbers should be taken with a gigantic grain of salt as I've just realized my laptop crunches matlab code at about ~1/10 the speed of a CAE computer. Go figure.

it's time to invest in new hardware.

dan

### Re: Timing Results, Kinematics Analysis

stepsize = 0.02

r - omegaBar ~145

r - p ~ 140

r - omegaBar ~145

r - p ~ 140

### Re: Timing Results, Kinematics Analysis

Hmm my results are kind of slow. But Matlab is occupying one thread at 100% during simulation. I should test it on a different machine as well.

r-p: 753s

r-w: 917s

stepsize:0.02

r-p: 753s

r-w: 917s

stepsize:0.02

### Re: Timing Results, Kinematics Analysis

Tolerance: 1.E-8

Step size: .02

Steps: 500

r-p: .577 [sec]

r-omega: .546 [sec]

Step size: .02

Steps: 500

r-p: .577 [sec]

r-omega: .546 [sec]

16 posts
• Page

**1**of**2**•**1**, 2Return to ME751 Spring 2010: Advanced Computational Multibody Dynamics

### Who is online

Users browsing this forum: No registered users and 1 guest