AnsweredAssumed Answered

Flow rate in gas nozzle not matching experimental results

Question asked by Gustavo Colheri Uchida on Sep 9, 2017
Latest reply on Sep 12, 2017 by Tg R2

Hello everyone,


I'm currently running simulations to determine the standard volumetric flowrate [Slpm] of a stove gas nozzle that we manufacture for a gas stove industry.


This is the gas nozzle:





The main objective is to compare the numerical results with our current measurements, since we are having some problems regarding flow tolerances with our client.



The last step of our drilling machine is a TEX-G4-VZ flowrate tester, that measure the standard volumetric flowrate of every piece during production.


We have colected a set of experimental results from the machine, and the average flowrate value was 2.333 [Slpm] with a pressure of 2.79 [kPa] (gauge). The ambient conditions at the time was 24.7ºC, 49% humidity, and pressure of 1017.9 hPa. At the time, we were using a 0.915mm drill. We have a set of drills with diameters ranging from 0.910 [mm], to 0.950 [mm], that we change during production to control the flowrate. The acceptable flowrate ranges from 2.300 [Slpm] to 2.400 [Slpm].


There are severals variables that changes the measurements values, we believe that roughness and hole diameter variations due to drill vibration are the major influencers.


In our machine test setup, the flowrate is currently measured in the opposite direction, because of space limitations.


Our client demands are that this model of gas nozzle have an air flow rate of 2.35 +/- 0.05 Slpm with 2.75 kPa gauge pressure, measured in the normal direction. The first conclusion we had with the simulations was that the flowrate changes with the direction of the flow for the same inlet pressure. This explain why some of our production batchs were approved during manufacturing, but reproved by the client when testing with the flow in the correct direction.


In our sets of simuations, we used ambient conditions of 25ºC, 50% humidity, 1013.25 hPa. The intlet airflow is 25ºC, 0% humidity, 2.75 kPa (gauge).


This is the simulation setup, in acordance with our test setup in the machine:setup.png


This is the mesh setup we used:



Here we have the velocity plot after 100 iterations (the flowrate converges by the 60th iteration), for both cases of flow (inverted and normal) with hole diameter of 0.92mm.



This is the Goal Plot during calculation

new mesh - INV - D92 - THETA 56 - H158.png


Here is a graphic with all our results, the blue line is derived from the simulations results with flow in the oposite direction (like our experimental setup), the orange lines is derived from the simulation results with normal direction flow (like our client setup). The green triangle is our experimental result derived from the average of 13 measurements.




The simulations results for flowrate are about 5.2% lower than experimental. We have changed several parametes, like roughness, meshing, pressure, and even simulated with the exact experimental conditions, but the numerical resultas are aways lower in the order of 5%.


What sould I do with the results? Apply a correction factor to equal with the experimental results, so we can use it to control our production?


Is there something wrong with the meshing and inlet/outlet parameters?


We have also checked the machine setup and the flow rate tester, there is no leakage, and the flow tester, that is borrowed from our client, has been calibrated this year.


Thanks to all,