Friction simulation
in [Matlab]
|
From: Math 50 on 10 Feb 2010 22:56 Hi, I'm designing a precision machine and need to figure out, before building the machine how friction will be affecting my system. I modelise my hole system on simulink but there is still the question of friction that is problematic. I dont really understand the parameter of the Coulomb & Viscous Friction block and dont even more understand how to fix it at real or representative values of my recirculating-ball slide (linear bearing). If somebody could help me to figure out the different parameter of the block and hoe to ajust it to represent my application, I would really appreciate it. Thank you Mathieu
From: Arnaud Miege on 11 Feb 2010 08:40 "Math 50" <jim(a)hotmail.com> wrote in message news:hkvv4l$kfg$1(a)fred.mathworks.com... > Hi, I'm designing a precision machine and need to figure out, before > building the machine how friction will be affecting my system. > I modelise my hole system on simulink but there is still the question of > friction that is problematic. I dont really understand the parameter of > the Coulomb & Viscous Friction block and dont even more understand how to > fix it at real or representative values of my recirculating-ball slide > (linear bearing). > > If somebody could help me to figure out the different parameter of the > block and hoe to ajust it to represent my application, I would really > appreciate it. > > Thank you > > Mathieu Which block are you referring to? Is it the Translational Friction block from Simscape? http://www.mathworks.com/access/helpdesk/help/toolbox/physmod/simscape/ref/translationalfriction.html The documentation explains pretty well what the different parameters are. You might also want to have a look at the following MATLAB Digest article and associated white paper + sample models: http://www.mathworks.com/company/newsletters/digest/2008/may/friction.html HTH, Arnaud
From: Math 50 on 11 Feb 2010 23:13 Hello, I'm using Simulink (straight), I dont have the license od Simscape. Yes, I want to simulate linear friction. As I see, the translation friction block for Simscape is quite clearer than the Coulomb&Viscous Friction block of Simulink . I guess that there's not equivalent block to the translation friction of Simscape in the normal toolbox of Simulink? Mathieu "Arnaud Miege" <arnaud.miege(a)nospam.mathworks.co.uk> wrote in message <hl11bp$ftk$1(a)fred.mathworks.com>... > > "Math 50" <jim(a)hotmail.com> wrote in message > news:hkvv4l$kfg$1(a)fred.mathworks.com... > > Hi, I'm designing a precision machine and need to figure out, before > > building the machine how friction will be affecting my system. > > I modelise my hole system on simulink but there is still the question of > > friction that is problematic. I dont really understand the parameter of > > the Coulomb & Viscous Friction block and dont even more understand how to > > fix it at real or representative values of my recirculating-ball slide > > (linear bearing). > > > > If somebody could help me to figure out the different parameter of the > > block and hoe to ajust it to represent my application, I would really > > appreciate it. > > > > Thank you > > > > Mathieu > > Which block are you referring to? Is it the Translational Friction block > from Simscape? > http://www.mathworks.com/access/helpdesk/help/toolbox/physmod/simscape/ref/translationalfriction.html > > The documentation explains pretty well what the different parameters are. > You might also want to have a look at the following MATLAB Digest article > and associated white paper + sample models: > http://www.mathworks.com/company/newsletters/digest/2008/may/friction.html > > HTH, > > Arnaud >
From: Arnaud Miege on 16 Feb 2010 05:06 "Math 50" <jim(a)hotmail.com> wrote in message <hl2kgi$l9k$1(a)fred.mathworks.com>... > Hello, > > I'm using Simulink (straight), I dont have the license od Simscape. Yes, I want to simulate linear friction. As I see, the translation friction block for Simscape is quite clearer than the Coulomb&Viscous Friction block of Simulink . I guess that there's not equivalent block to the translation friction of Simscape in the normal toolbox of Simulink? > > Mathieu > I'm confused. There is no Coulomb & Viscous friction block in Simulink. This must be a custom block that somebody else has created. You could maybe create your own translational friction block in basic Simulink or with an Embedded MATLAB block, using the equations from the Simscape block. Good luck, Arnaud
From: Arnaud Miege on 16 Feb 2010 09:28 "Arnaud Miege" <arnaud.miege(a)nospam.mathworks.co.uk> wrote in message news:hldqmd$lf2$1(a)fred.mathworks.com... > "Math 50" <jim(a)hotmail.com> wrote in message > <hl2kgi$l9k$1(a)fred.mathworks.com>... >> Hello, I'm using Simulink (straight), I dont have the license od >> Simscape. Yes, I want to simulate linear friction. As I see, the >> translation friction block for Simscape is quite clearer than the >> Coulomb&Viscous Friction block of Simulink . I guess that there's not >> equivalent block to the translation friction of Simscape in the normal >> toolbox of Simulink? >> >> Mathieu >> > > I'm confused. There is no Coulomb & Viscous friction block in Simulink. > This must be a custom block that somebody else has created. > > You could maybe create your own translational friction block in basic > Simulink or with an Embedded MATLAB block, using the equations from the > Simscape block. > > Good luck, > > Arnaud > My mistake, I have found the block you were referring to: http://www.mathworks.com/access/helpdesk/help/toolbox/simulink/slref/coulombandviscousfriction.html If you look at the shape of the curve and compare it to that of the Simscape friction block, you'll notice that it is essentially the same, except for two main differences: - there is no Stribeck friction in the Simulink version - the slope around zero velocity is infinite, whereas in Simscape it is finite. The "Coulomb friction force" parameter in the Simscape block is therefore equivalent to the "Coulomb friction value (offset)" parameter in the Simulink version, and the "viscous friction coefficient" parameter in the Simscape block is equivalent to the "coefficient of viscous friction (gain)" parameter in the Simulink version. As mentioned, you could try implementing your own version using Simulink blocks or Embedded MATLAB. Here's an example, which I have tested function f = trans_friction(v,Coulomb_F, break_away_F, visc_fric_coeff, trans_coeff, vel_thr) %#eml brkwy_frc_th = visc_fric_coeff * vel_thr + Coulomb_F + (break_away_F - Coulomb_F) * exp(-trans_coeff * vel_thr); if (abs(v) <= vel_thr) % Linear region f = brkwy_frc_th * v / vel_thr; elseif v > 0 f = visc_fric_coeff * v + Coulomb_F + (break_away_F - Coulomb_F) * exp(-trans_coeff * v); else f = visc_fric_coeff * v - Coulomb_F - (break_away_F - Coulomb_F) * exp(-trans_coeff * abs(v)); end where I have defined Coulomb_F, break_away_f, visc_fric_coeff, trans_coeff, and vel_thr to be parameters rather than inputs to the Embedded MATLAB function. HTH, Arnaud
|
Pages: 1 Prev: Least Squares fit non-linear 3d plot Next: return values from boolean |