FacebookTwitterDiggStumbleuponGoogle BookmarksRedditTechnoratiLinkedin

{[Mass Effect]} Presentation of the Mass-Flow Calorimeter

Written by Mathieu Valat on .

After almost a year of work, the Mass-Flow Calorimeter (MFC) system is finally ready. It's been a great challenge and it was far from simple to make.

In principle it is meant to precisely measure the quantity of energy given by a sample. And because we are aiming at precision here, it was necessary to leverage painstakingly multiple expertises. Moreover, the bench will be used for many configurations and for other experiments where we are shooting for great precision. There is even an extension for another experiment to be run at the same time.

There have been one key help in the design, mounting, configuring and setting-up this big bench : Jean-Paul Biberian. His most valuable support was offered throughout the critics that are absolutely necessary for a scientific work to be done properly. His casualness of sharing his scientific expertise make the ride so much rewarding, I am so thankful!

Here is a detailed presentation of this bench.

I am taking advantage of this blogpost to thank our supporters, you, who helped us and enabled us to make this system.

Go Open-Science!

 

Mathieu Valat

Comments   

 
0 #4 Mathieu Valat 2015-09-02 09:50
Quoting Marissa Little:

Are you recording room temperature near the calorimeter?

Of course!

Quoting Marissa Little:
We have found correlation between room temp and the outlet water temperature despite heroic efforts (temperature controlled outer box, insulated inner box, tightly controlled water temperatures, etc). We've seen this in several systems. It is very minor but sometimes sheds light on interesting artifacts that show up in these type of tests. Better to have the room temp data on hand from the beginning in my opinion!


Indeed Marissa. I see also a 0.1°C variation of the water inlet with static water. It goes down to 0.02°C when water flow is on. I have been tweaking a lot the regulation too. These efforts are huge also, and I am not yet finished with that. :cry:

Flow is set to 143ml/min so 0.1°C is equivalent to 1W.

But on the positive side, I see these variation only when the change of room temperature is above 1.5°C/hour. This is less likely to happen when I turn off the Celani cells next to it. They add 100W permanently to the room.
If it is below this threshold, the PID compensate that easily.

My objective is to reach 100mW resolution. But that will be difficult, I know it. :)

Quoting Marissa Little:
We found the positive-displacement pump (from FMI) did not produce stable enough flow rates so removed the included motor for a precision motor. We still found that we could see flow changes due to fluctuations of line frequency! So we ran the motor with a precision-frequency supply. The only way we knew all of these fluctuations were happening was with a custom-made flow meter and not simply relying on the pump setting. Just something to think about.


I am very much aware of this FMI pump limitation. But due to money constrain, I am not able to change that part for a digital mass-flow controller. I am in contact with Bronkhorst NL, see if they can help me our with that. Depending on them, I shall request details of your solution...

Quoting Marissa Little:
In the early days of our calorimeter, we had odd results that took us a long time to track down. Small bubbles were accumulating somewhere in the line until the bubble was large enough to break free and flow out. This caused a sudden change not only to the mass flow for a brief time, but to the "wetting" along the copper tube and a transient temperature change of the outlet water despite no change to the actual power being delivered to the water. This was resolved quickly with the addition of bubble traps.


Good to know, I shall take that precaution too.

Quoting Marissa Little:
My biggest piece of advice is to run lots and lots of calibrations/control runs! Try putting resistive heaters at various locations in the chamber - dead center, at one end, close to one side, etc. Location sensitivity can be a real issue.


That is also in the plan.

Quoting Marissa Little:
Anyway, it looks great and you've done a good job!


Thanks!
Quote
 
 
0 #3 Marissa Little 2015-08-25 19:32
Amazing work Mathieu! Just a couple comments/questions:

Are you recording room temperature near the calorimeter? We have food correlation between room temp and the outlet water temperature despite heroic efforts (temperature controlled outer box, insulated inner box, tightly controlled water temperatures, etc). We've seen this in several systems. It is very minor but sometimes sheds light on interesting artifacts that show up in these type of tests. Better to have the room temp data on hand from the beginning in my opinion!

We found the positive-displa cement pump (from FMI) did not produce stable enough flow rates so removed the included motor for a precision motor. We still found that we could see flow changes due to fluctuations of line frequency! So we ran the motor with a precision-frequ ency supply. The only way we knew all of these fluctuations were happening was with a custom-made flow meter and not simply relying on the pump setting. Just something to think about.

In the early days of our calorimeter, we had odd results that took us a long time to track down. Small bubbles were accumulating somewhere in the line until the bubble was large enough to break free and flow out. This caused a sudden change not only to the mass flow for a brief time, but to the "wetting" along the copper tube and a transient temperature change of the outlet water despite no change to the actual power being delivered to the water. This was resolved quickly with the addition of bubble traps.

My biggest piece of advice is to run lots and lots of calibrations/co ntrol runs! Try putting resistive heaters at various locations in the chamber - dead center, at one end, close to one side, etc. Location sensitivity can be a real issue.

Anyway, it looks great and you've done a good job!
Quote
 
 
0 #2 Sanjeev 2015-08-23 10:19
Excellent work !
The excess heat has no place to hide with such a setup.
Quote
 
 
0 #1 bob 2015-08-21 12:55
Mathieu: Very impressive effort.

I think your apparatus offers a great opportunity to compare and calibrate other calorimeter ideas. One that I believe should be considered is the conduction calorimeter http://www.quantumheat.org/index.php/it/follow/follow-2/364-mfmp-followers-propose-a-new-calorimeter

A very simple conduction calorimeter could be constructed as two concentric Al tubes separated by a small air gap (the insulation layer). Temp sensors could go on ID of outer tube and OD of inner tube yielding delta T across the air gap.

To test this calorimeter against your mass flow one it would be a matter of using a joule heater coil inside the conduction calorimeter inner tube and then inserting the whole assembly inside your mass flow calorimeter. Input power to the joule heater should match the heat loss in your mass flow calorimeter and should yield a linear delta T vs input power on the conduction calorimeter.

In any case keep up the good work and I look forward to results from the new device.
Quote
 

Add comment


Here is your generous contributions so far towards our $500,000 target, thanks everyone! : $45,020   Please Donate
See the current state of our booked costs here