Tank Volume in PLC and HMI

[Bullzi]

Norm,
You are correct the 3rd order equation was much closer. It is only about 0.5% off at 131". That is much better. Actually it is much better through out the curve. I just figured that the higher the order the better the curve. Thanks Norm!!

Lance,
Thanks for the info on how to look at the curve much better. I figured there would be some errors especially at the ends of the curve. I just couldn't see it until you told me about it.

A 3rd order equation looks like it will do the job within the accuracy I need. A big thanks to Tom for showing us all how to use this feature in Excel. It is very powerful!! And a big thanks to you Lance for doing a lot of leg work on this.

 

[ndzied1]

I try to watch out for big numbers multiplied by small numbers.

In the 4th order you have a very small coefficient for x^4.

 

[dlweber]

+1 Tom
+1 Excel

I have done similar things when I was doing Digital filter design. It really came in handy when trying to figure out what was wrong when using fixed point programming.

 

[Luke_Mages]

I only scanned through this but has anyone mentioned using simple piecewise function?

I have several AOIs I wrote for different shaped tanks that do this.

 

[Bullzi]

I only scanned through this but has anyone mentioned using simple piecewise function?

I have several AOIs I wrote for different shaped tanks that do this.

Luke,
Would you mind expanding on this?

 

[Luke_Mages]

Luke,
Would you mind expanding on this?

Sure thing...

The Anderson level sensors we typically use measure inches of column...so from the tank drawings or from actual measurements create piecewise functions that calculate volume based on the shape of the tank. For example, a tank I calibrated recently had a dished bottom. The dish was 4 inches deep and had a radius of 128". so for the first 4" the volume is calculated using the volume of a dome with radius of 128". above 4" volume is calculated using the volume of a cylinder. Then there is also a specific gravity offset to compensate for different liquids that might be put into the tank.

 

[Lancie1]

Lance,
Thanks for the info on how to look at the curve much better. I figured there would be some errors especially at the ends of the curve.

Comparing your DIP Chart to the calculated gallons for an ideal horizontal cylindrical tank with flat ends, it appears that your tank has more of an eliptical cross-section instead of a perfect circle. See the attached spreadsheet. I believe that this is typical for real-world tanks, and one reason why a measured Dip Chart is usually more accurate than a calculated volume based on an ideal shape.