x,y,z and time plot with temperature

mande01

Does anyone know of a simple method for plotting temperature in 4 dimensions.

As in i have a vessel which i am temperature mapping.
I have probes in 3 dimensional locations and I want to represent them in a graphic over time.

I know with ansys or pro eng this can be done.
But I looking for a simple solution, if it exsts, other than buying ansys or pro eng to do this one simple task.

Thank you,
Derry

professore

Have a look at this : http://mayavi.sourceforge.net

Never used it myself but looks good

coolwings

4DFieldPro will very likely do what you want.
Look here: http://field.hypermart.net/

Turbulent Bill

I think MATLAB would do a good job on this (if you have access to a copy). You could create a 3D plot to represent the measurement positions, with a temperature colourmap to represent the temperature at each position. Do a separate plot for each time interval, then stitch them together in a video/animation.

mande01

Thank you one and all for your responses.
I was really surprised with the speed and quality, only just after joining boards.ie
I'm about to check out the two first options, matlab I know can do this but the problem is cost.

The problem in more detail is as follows if more information is out there:
I have a chamber, ~ 4 m x 3 m x 3 m
It operates in a temperature range from -60 deg c to + 60 deg C for a production cycle
And up to 130 deg C for a prep cycle
A specific cycle last 24 hours, the temp varies from -60 to +60,
During this cycle I have 60 thermocouples (T/C) spread throughout the chamber
5 T/C at 12 levels (four corners and centre)

I want to be able plot these in a graphical sense which will show the temp vary as time goes on.

Ideally being able to zoom in to an area and slide a bar to vary time.

I now go to get the demos of the applications

Thanks again,
Derry

P.S. this is a lyophiliser in a pharma and I'm trying to find out why its not performing correctly.

P.S.S the sample rate is every 30 sec to 1 min so lots of still pictures won't be practical

franman

Are you just looking for hot and cold spots in the lyo? Why cant you just use excel for this? Take a record of what T/C is at what location. Output the temperature data to excel and the graph temp vs time.
When during the cycle is it failing? Are you passing the pressure test with the TCs in place?

mande01

Hello franman,
The lyo hasn't failed what we are say was drift, and I want to put in place a simple way for people to visualise the data that gets generate from the mapping exercise.
(just to note the vendor response was ????, its hasn't failed what's the problem?, I want anyone to be able to see the data and immediately say "that's not right".)

The problem is the understanding of how the system is drifting and ideally where.
Being able to visualise all the data points at once would make investigations, and decision making, much easier.

In a recent issue, the solution was simple, the problem was obvious to those who knew that data but transferring this information to people who were not actively in involved was very difficult.


But any way. As I've being writing this out, I've come to the conclusion that I'm going to go the Matlab root. i know it can do it so I'll just get on to Matlab for a trail and try to get it done in the 30 days. Then see if we can buy it.

On the example solutions above I've tried them both but I don't know how to generate the data sets for them.
Any one have experience of this? Or know where a good tutorial would be?
Thanks,
Derry

Cheeble

Are you trying to make it too complicated? I had a similar problem some years ago and we just did this -



The legend colours were assigned according to position so areas and trends would show up.

Cheeble-eers

mande01

In essence I am trying to make it too complicated!

We have been down this road with charts and graphs.
Moved from that specific format and now present the data normalised against the set point. i.e. at -50 set point subtract -50 and just plot the deviation from 0.
This helps compare all set point variations against each other. As in it shows weather or not it is easier to control at 20 deg C than at -50 deg C. which it obviously is.

The chart above does not show you, not obviously, that say there is a cold area in the top right hand corner, across the top 2 shelves. It involves trying to transfer the data from the 2d representation above to a 3d view in your head.
Most people who are involved in this type of work can do this quite easily as they are used to reviewing data but others ,who do not have similar abstract visualisation abilities, find it quite difficult to identify with what is going on.

I want a 3d picture that shows the locations in the chamber varying according to there temperature over time. Ideally by varying the time manually with a sliding scale, or roll of a mouse wheel!

I waiting on a trail CD from matlab and I'll report back.
But if someone out there has does something similar in any environment I would love to hear about it.

Thanks,
Derry

backboiler

Since cost is a factor, consider Octave, which is similar to Matlab.
http://www.gnu.org/software/octave/
Simple Matlab .m scripts will work with minor modifications in Octave.
I've only ever used it in Linux or in Windows under Cygwin but there seems to be a standalone binary for Windows on the site too as well as other platforms.