Altair Flux Motor Model Couldn't Recognize on PSIM

abdin ulin

Dear Expert

I just run example motor design suite flux motor, i have build my own motor data and successfully load the custom motor data "flux2.mat", PSIM prom graph Cannot find graph output axis : LOSS_IRON in graph

please help how to generat ".mat" file properly hence PSIM could recognize this motor model

Thank you

BR

Abdin

NikosDimitrakopoulos

Dear Abdin,

This is a known incompatibility issue with FluxMotor 2023 and 2023.1 -> expected to be solved in v2024. For your reference, this is documented in this discussion:

https://community.altair.com/community?id=community_question&sys_id=4c474db31b26f554507ca6442a4bcb62&view_source=searchResult

The issue is attributed to a difference in the name of "iron losses" and "torque" headers.
With FluxMotor 2022.3 it was LOSS_IRON
with FluxMotor 2023, it is LOSS_IRON_STATOR.

______________________________________________________________________

With FluxMotor 2022.3 it was TORQUE
with FluxMotor 2023, it is ELEMENT_TORQUE_ELECTROMAG.

_______________________________________________________________________

As a workaround, you can open the .mat file exported from FluxMotor using Compose and the MAT Explorer tool. From there, you can rename the headers accordingly. Once modified, you should save the altered file using the save('path/name.mat') command.

Or you can always use FluxMotor 2022.3 to export.

Hope this helps!

BR,
Nikos Dimitrakopoulos

abdin ulin

Nikos Dimitrakopoulos said:

Dear Abdin,

This is a known incompatibility issue with FluxMotor 2023 and 2023.1 -> expected to be solved in v2024. For your reference, this is documented in this discussion:

https://community.altair.com/community?id=community_question&sys_id=4c474db31b26f554507ca6442a4bcb62&view_source=searchResult

The issue is attributed to a difference in the name of "iron losses" and "torque" headers.
With FluxMotor 2022.3 it was LOSS_IRON
with FluxMotor 2023, it is LOSS_IRON_STATOR.

______________________________________________________________________

With FluxMotor 2022.3 it was TORQUE
with FluxMotor 2023, it is ELEMENT_TORQUE_ELECTROMAG.

_______________________________________________________________________

As a workaround, you can open the .mat file exported from FluxMotor using Compose and the MAT Explorer tool. From there, you can rename the headers accordingly. Once modified, you should save the altered file using the save('path/name.mat') command.

Or you can always use FluxMotor 2022.3 to export.

Hope this helps!

BR,
Nikos Dimitrakopoulos

Dear Nikos

Thank you very much for your respon, yeah a day later i just figuring out that PSIM failed to load table LOSS_IRON and also TORQUE, a simple fix rename table header would solve the problem as shown below and save then replace / overwrite, now PSIM could load the .mat file 

 

 now another anomalies happen, even i have input every parameter based on this tutorial

I have click "update parameter" file button, seems Torque speed curve generated by PSIM wouldn't load a correct value, the max torque still at 200 Nm even based on .mat file my motor design a 309 Nm

I have double checked manually on PSIM table, but PSIM torque speed curve still not updated yet

another question rised, is that normal almost column have 100 point index except for LOSS_IRON and LOSS_TOTAL that have 1000 point index, i affraid there will missmatch calculation on PSIM

Your help would be appreciated

BR
Abdin

NikosDimitrakopoulos

abdin ulin said:

Dear Nikos

Thank you very much for your respon, yeah a day later i just figuring out that PSIM failed to load table LOSS_IRON and also TORQUE, a simple fix rename table header would solve the problem as shown below and save then replace / overwrite, now PSIM could load the .mat file 

 

 now another anomalies happen, even i have input every parameter based on this tutorial

I have click "update parameter" file button, seems Torque speed curve generated by PSIM wouldn't load a correct value, the max torque still at 200 Nm even based on .mat file my motor design a 309 Nm

I have double checked manually on PSIM table, but PSIM torque speed curve still not updated yet

another question rised, is that normal almost column have 100 point index except for LOSS_IRON and LOSS_TOTAL that have 1000 point index, i affraid there will missmatch calculation on PSIM

Your help would be appreciated

BR
Abdin

Dear Abdin,

I am happy that you were able to load the .mat. Your following questions are also good points. Let me directly get to them:

1. For the Speed-Torque Curve tool, there are parameters that can limit the motor's output.  Most often than not, DC Bus voltage and Inverter Current are the culprits:

The Speed-Torque curve will give you the maximum op. points of the system based on your input voltage and control limits. I suggest that you also read this tutorial:

Are you sure that the input voltage is high enough to drive the motor with the given back EMF? Also, maybe the control detects a current limit (Is_max) and limits the amount of inverter current. What was the current and voltage that was used to create the LuTs in Flux? You can experiment with those values to see if the results change.

2. When it comes to the 100 vs 1000 points issue, I suspect that you used 10 Id points, 10 Iq points and 10 speed points to calculate the LuTs. The Id-Iq combination of points is 10*10 = 100 different operating points. Torque and Copper losses rely on the Iq-Id combination so this is the reason why 'Torque' and 'Loss_Joule' columns have 100 points. On the other hand, the equations determining the 'Loss_Iron' column also rely on the different speed points along with the Iq-Id combinations. Remember that iron losses are highly affected by the motor speed. That means that we need to have 100 different Iq-Id combinations for each one of the 10 different speed points. That makes a total of 100*10=1000 'Iron Losses' combinations.

Hope this helps!

Best regards,

Nikos Dimitrakopoulos

 

 

abdin ulin

Nikos Dimitrakopoulos said:

Dear Abdin,

I am happy that you were able to load the .mat. Your following questions are also good points. Let me directly get to them:

1. For the Speed-Torque Curve tool, there are parameters that can limit the motor's output.  Most often than not, DC Bus voltage and Inverter Current are the culprits:

The Speed-Torque curve will give you the maximum op. points of the system based on your input voltage and control limits. I suggest that you also read this tutorial:

Are you sure that the input voltage is high enough to drive the motor with the given back EMF? Also, maybe the control detects a current limit (Is_max) and limits the amount of inverter current. What was the current and voltage that was used to create the LuTs in Flux? You can experiment with those values to see if the results change.

2. When it comes to the 100 vs 1000 points issue, I suspect that you used 10 Id points, 10 Iq points and 10 speed points to calculate the LuTs. The Id-Iq combination of points is 10*10 = 100 different operating points. Torque and Copper losses rely on the Iq-Id combination so this is the reason why 'Torque' and 'Loss_Joule' columns have 100 points. On the other hand, the equations determining the 'Loss_Iron' column also rely on the different speed points along with the Iq-Id combinations. Remember that iron losses are highly affected by the motor speed. That means that we need to have 100 different Iq-Id combinations for each one of the 10 different speed points. That makes a total of 100*10=1000 'Iron Losses' combinations.

Hope this helps!

Best regards,

Nikos Dimitrakopoulos

 

 

Dear Nikos,

Thank you very much for immediate answer and help, very very detail, this topic just get more challenging, lets get back to the basic of PSIM example, lets have the same reference built in PSIM example, there are question i will ask

lets say we have opened Motor Control Design suite - PMSM (IPM) Drive (FluxMotor) as shown below

then we have full example project shown below

and then we open Fluxmotor nissan leaf project as shown, based on Machine performance - base speed point, in general data table Mechanical torque stated 219.862 Nm

and we open simview and find "NissanLeafLikeMotor_LUT_FluxMotor.mat" and load to simview windows

I suspect this Torque data its very different from Torque vs Speed Curve and actually just above Mechanical torque stated 219.862 Nm, as we remember the Torque data shown below just increase and decrease relative to the index value just like sine wave, I suspect the Torque data its from Torque vs Angle or Torque vs something else instead from Torque vs speed.

so that there is no any direct correlation between Torque data shown above with Torque vs Speed characteristic shown below 

I just barely understood how the way PSIM work in this case

I suspect not all data in .mat file was used by flux motor block, i guest LOSS_IRON and LOSS_Joule was exactly inherently used due to we have probing Ploss_copper & Ploss_Core, the Torque data not inherently picked from Torque table, instead PSIM would calculated based on Current, Ld, Lq, Flux d, Flux q in spesific IPM motor formula 

I just wonder its seem even i have good machine model, based on what i have design on Fluxmotor

not always have the same with Torque Speed Curve generated from PSIM (looks similiar now), but not looks similiar when i have used my custom motor model .mat

of course every parameter have correctly entered

I have attached in .rar my custom motor design if you want replicate or playing arround what i currently simulating that motor into PSIM environment

it just seem not quite right, 

The Question is :

" I have design very high Torque Motor, 411Nm (base speed torque) but when simulating in PSIM environment, by just applying 320Nm ish Torque, the motor just stall, not able to catch the speed reference set point, I suspect PSIM just failed to model my custom motor, the custom motor model just not have enough Torque capacity simulated at PSIM environment eventhough the Fluxmotor model succeed modeled at 411Nm base torque. "

Quite late night here, I very very appreciate your respon and answer and of course your patient answering my question lately

Thank you very much

BR

Abdin

NikosDimitrakopoulos

abdin ulin said:

Dear Nikos,

Thank you very much for immediate answer and help, very very detail, this topic just get more challenging, lets get back to the basic of PSIM example, lets have the same reference built in PSIM example, there are question i will ask

lets say we have opened Motor Control Design suite - PMSM (IPM) Drive (FluxMotor) as shown below

then we have full example project shown below

and then we open Fluxmotor nissan leaf project as shown, based on Machine performance - base speed point, in general data table Mechanical torque stated 219.862 Nm

and we open simview and find "NissanLeafLikeMotor_LUT_FluxMotor.mat" and load to simview windows

I suspect this Torque data its very different from Torque vs Speed Curve and actually just above Mechanical torque stated 219.862 Nm, as we remember the Torque data shown below just increase and decrease relative to the index value just like sine wave, I suspect the Torque data its from Torque vs Angle or Torque vs something else instead from Torque vs speed.

so that there is no any direct correlation between Torque data shown above with Torque vs Speed characteristic shown below 

I just barely understood how the way PSIM work in this case

I suspect not all data in .mat file was used by flux motor block, i guest LOSS_IRON and LOSS_Joule was exactly inherently used due to we have probing Ploss_copper & Ploss_Core, the Torque data not inherently picked from Torque table, instead PSIM would calculated based on Current, Ld, Lq, Flux d, Flux q in spesific IPM motor formula 

I just wonder its seem even i have good machine model, based on what i have design on Fluxmotor

not always have the same with Torque Speed Curve generated from PSIM (looks similiar now), but not looks similiar when i have used my custom motor model .mat

of course every parameter have correctly entered

I have attached in .rar my custom motor design if you want replicate or playing arround what i currently simulating that motor into PSIM environment

it just seem not quite right, 

The Question is :

" I have design very high Torque Motor, 411Nm (base speed torque) but when simulating in PSIM environment, by just applying 320Nm ish Torque, the motor just stall, not able to catch the speed reference set point, I suspect PSIM just failed to model my custom motor, the custom motor model just not have enough Torque capacity simulated at PSIM environment eventhough the Fluxmotor model succeed modeled at 411Nm base torque. "

Quite late night here, I very very appreciate your respon and answer and of course your patient answering my question lately

Thank you very much

BR

Abdin

Dear Abdin,

Could you let me know what version of PSIM and FLux are you using? Do you face the same issue if you export from Flux 2022.3 instead of renaming the torque column?

Also what is the type of license you have?

BR,

Nikos