ivar_kjelberg

51 Reputation

3 Badges

14 years, 103 days
My current post is "Senior Expert" with 30+ years of experience in mechatronics developments in the frame of complex, multidisciplinary, high tech "systems engineering" activities. Previous posts were "Senior Project Manager, Section Head, Project Manager, Quality Engineer, R&D Engineer. I work in a retained profit company, a "High Tech", R&D Institution, closely linked to the EPFL (Ecole Polytechnique Federale de Lausanne), we are located in Neuchatel Switzerland. We use advanced tools such as Maple, MapleSim and COMSOL Multiphyiscs for our developments and for our Ph.D. training activities.

MaplePrimes Activity


These are replies submitted by ivar_kjelberg

Hello again

Indeed, I see there is no way to input a "." variable in the 2D math display of MapleSim, here the name fluidMedium.cp is automatically transformed into a matrix multiply, I suspect.

So, I'll rewrite my full model, it's gonna take a day, and I have already lost 2 now, but I have learned a bit more about the internals of MapleSim. It's just that my boss does not pay me for that, nor for lost time, so I must hurry up to get this model running ...

Thanks again for the help
and lets hope the developpers speeds up a bit, and carefully test out the software, also for subsystems, next time :)

Sincerely

Ivar

@OrangVahid 

Hi and thanks,

indeed that could explain the "p" issue. I see that the problems then comes from a somewhat poorly defined Modelica item as implemented in MapleSim: "Modelica.Thermal.FluidHeatFlow.Interfaces.FlowPort_..." that does not refer correctly to the "medium=fluidMedium" used in MapleSim.

Then you also confirm what I suspected that the "nice" math display of MapleSim can be very misleading here.

But, what suprises me is that this should be known and easily corrected and we are already at MaplSim 2017 v3 (still my first excursion in the thermal fluid domain though ;) Hopefully its improved in v.2018 to come I hope soon

OK I'll adapt then this way. But not using the subsystems, is a real pain for my "heavy" model with six quite complex  subitems ...

Thanks for your help

Sincerely

Ivar

Hello again

Here is a simple model to illustrate the case, this one runs as is: (see file link below)

But, select the IP1 pipe and the wires and make a subsystem => run =>

error of type a) above

Fix select subsystem entry frame and change the "1" to rho C[p] C[v] lambda and nu and it will run further =>

error type b) above

"p" is missing, just as if the pressure is not transported inside the block ...

Here I'm stuck !

I'm running a 2017.3 version, no specialitites on Win7 64 up to date

file here:  Ilustrate_Cp_p_error2017_3.msim

Thanks for any suggestions

Ivar

So this one was quickly fixed :)

Thanks for that Ivar

Hi Graham

Well what about considering adding start & stop times for all graphs, there is enough room on your plot tab. And leave the default simulation sart & stop values as "default", you do it many places in the blocks.

 

Pls consider for future releases

Sincerely

Ivar

 

PS, I will come with a few other suggestions too then, I have found several things that would make life so much easier for us users (at least those having a boss with a stopwatch looking over our shoulders ;)

Hi Graham

Well what about considering adding start & stop times for all graphs, there is enough room on your plot tab. And leave the default simulation sart & stop values as "default", you do it many places in the blocks.

 

Pls consider for future releases

Sincerely

Ivar

 

PS, I will come with a few other suggestions too then, I have found several things that would make life so much easier for us users (at least those having a boss with a stopwatch looking over our shoulders ;)

Hi Graham

Good to hear you understand my demand, and nice to learn that you consider it for future releases, I have just to be patient then ;)

Sincerely

Ivar

Hi Graham

Good to hear you understand my demand, and nice to learn that you consider it for future releases, I have just to be patient then ;)

Sincerely

Ivar

@Joe Riel 

Thnaks for the precisions Joe

I will take some more time to study the Modelica outputs then, but for me most of my simulations are a mixture of algebraic equations and block build items (the later typically for a mechanical geometry/structure). But everything can not easily be put into blocks, I need the equation inputs. I.e I have a non-linear spring (linking two parts of my structure) that depends on the local position of my structure. This implies acting on the spring constant k(x), I do not see how to get around only by adding forces ;)

Well more work for me, it will not be as easy as hoped, but lets see the nice side, I'll learn something new ;)

Thanks again

Ivar

@Joe Riel 

Thnaks for the precisions Joe

I will take some more time to study the Modelica outputs then, but for me most of my simulations are a mixture of algebraic equations and block build items (the later typically for a mechanical geometry/structure). But everything can not easily be put into blocks, I need the equation inputs. I.e I have a non-linear spring (linking two parts of my structure) that depends on the local position of my structure. This implies acting on the spring constant k(x), I do not see how to get around only by adding forces ;)

Well more work for me, it will not be as easy as hoped, but lets see the nice side, I'll learn something new ;)

Thanks again

Ivar

Hi Joe

Well then I'm missing some of the interest in MapleSim, I need to be able to loop my variables such that i.e I have a spring constant "k" that changes with position, or that I adapt a value as function of time.

for me the variables x,y,z and t should be adressable in my variables. For x,y,z I need to ba able to use my own references too, I can set a relative position sensor, but then I ne to feed it back into a variable of another block

Ivar

Hi Joe

Well then I'm missing some of the interest in MapleSim, I need to be able to loop my variables such that i.e I have a spring constant "k" that changes with position, or that I adapt a value as function of time.

for me the variables x,y,z and t should be adressable in my variables. For x,y,z I need to ba able to use my own references too, I can set a relative position sensor, but then I ne to feed it back into a variable of another block

Ivar

Thanks,

this work-around looks great, it will save me hours of clicking

I hope Maple will add soon a global plot time axis selection in their plot window, as often one want to skip the first 1-2 periods to leave your system settle (and to avoid the spikes that scales your graphs out of usefull range)

So long

ivar

Thanks,

this work-around looks great, it will save me hours of clicking

I hope Maple will add soon a global plot time axis selection in their plot window, as often one want to skip the first 1-2 periods to leave your system settle (and to avoid the spikes that scales your graphs out of usefull range)

So long

ivar

And if I have a piecewise function (continuous derivable to the "jerck"), depending on a several line internal function, how can I get that into a block such to replace the simpel "real-signal Sine" block ?

Because none of the templates are right fo that, I have a function not a series of data points, nor a PDE

 

Thanks in advance

Ivar

1 2 Page 1 of 2