cancel
Showing results for
Did you mean:

Solved!
Go to solution

This is probably quite simple to do, but as yet I have only got part way there, so I could use some help!

Basically I have some shaft torque and speed datalog files, from which I like to extract a table of number of shaft revolutions at discrete torque levels for use in CAE fatigue analysis, like this:

423 revs @ 1000 Nm

569 revs @ 1200 Nm

143 revs @ 1400 Nm

The "quantize" script gets me part way there - this can reduce the torque curve to 200 Nm steps, and it's easy enough to integrate the speed to give the cumulative number of revolutions, but now I need to count the number of revolutions at each quantized torque step, and output the results.

There must be a way to do this with all the tools available in DIAdem, but being a new user, I'm struggling to find the correct method.

Any help would be appreciated, thanks.

Message 1 of 6
(6,003 Views)
Solution
Accepted by topic author JohnGrove

Hi JohnGrove,

It sounds to me like you need to use one of the "CLassification" ANALYSIS functions to count the number of revolutions within certain predefined torque buckets.  You can specify exactly which torque conditions to use for the counting in the "Classification" ANALYSIS function.

So the main question is how to create a Revolutions channel from the data you're starting with.  It's not clear to me that this can be done with just a Speed channel.  Do you have a Tachometer channel, or do you know the circumferance of the wheel and that it is contacting the pavement at all times?  Once we get the Tachometer channel you'll be home-free.

National Instruments

Message 2 of 6
(5,757 Views)

Im not intending to steal this thread but I have a quesiton very similar to this...

I have some RPM and Load data in addition to Fuel Consumption data. Attached are some pictures of an example an individual from Germany has done along these same lines but I cannot seem to replicate the process.

My goal is to generate a 3D plot with Engine Speed (RPM) and Engine Torque (%) on the x and y - axis respecitively and % Fuel used on the z-axis. I've been playing around with the Compound Classification function in the ANALYSIS tab.

Can someone point me in the right direciton regarding this?

Thanks,

~Nate

The first attachment is the example from the German individual

The second attachment is my current setup

Message 3 of 6
(5,677 Views)

Hi Nate,

It looks like the easiest way to get to that graph would be to start in VIEW with a standard Contour Plot with the top and right graphs that show the 2D slices for the current position of the VIEW crosshair cursor on the Contour Plot.  Check out the example called "Contour Display in DIAdem VIEW", for instance.  Once you have that set up in VIEW, then you can transfer that sheet from VIEW ==> REPORT to have an excellent starting point for the REPORT template.  You will then need to change the two 2D graphs from line display to vertical/horizontal bar display, and you will need to change the 3D graph from characteristic diagram display to 2D matrix display with the width and offsets all set to 0.

You'll also need to replace the actual channels displayed in the two 2D graphs in VIEW with the real channels from the classification calculations.

Are your troubles more on the calculation side or the REPORT display side?

National Instruments

Message 4 of 6
(5,666 Views)

Thanks for the response. Populating the graphs is pretty straight forward, the trouble I'm having is how to bin the data into a collective. The example I was given utilized the compound classification tool and I can't seem to get the functionality out of it that the example shows. I'll play around with it some more today but if you could give a few pointers regarding that it would be much appreciated.

Thanks,

~Nate

Message 5 of 6
(5,643 Views)

Bit of a late reply, sorry, but I just picked this up again as I have a couple of huge batches of data ready for analysis. Anyway, I was pleased to find that once I'd got my head around it, your suggestion of classification analysis was spot on.

All I had to do was multiply the speed by the datalog time period to give me the distance travelled per data point and then divide that by the tyre circumference to give the number of revolutions per data point. Then the single reducing classification analysis with the summing option checked gave me exactly what I needed.

Regards,

John.

Message 6 of 6
(5,374 Views)