BUG: velocities computed from stiffnessTensors from ODF with uniform components is wrong

[kilir]

Hi Ralf, when we have an ODF which contains mixtures with uniformComponents, the computed stiffnessTensors, respectively following results e.g. velocities can be higher or lower then single crystal values. I don't think physics allows this. Does it?

CS= crystalSymmetry('-3m', [5.323 5.323 7.272], 'X||a*', 'Y||b', 'Z||c');
rho= 2.5155; %g/cm³
N=[[229.08    89.04       13.56          0         0        0];...
   [89.04    229.08       13.56          0         0        0];...
   [13.56     13.56       45.84          0         0        0];...
   [0             0           0      12.77         0        0];...
   [0             0           0          0     12.77        0];...
   [0             0           0          0         0    70.02]];
E=stiffnessTensor(N,CS)

% vp single crystal
VP_SC = velocity(E,'density',rho);

% vp fibre
cfib = fibreODF(Miller(0,0,0,1,CS),yvector,'halfwidth',30*degree)

[CV, CR, CH] = calcTensor(cfib,E);
vp_Fib_V = velocity(CV,'density',rho);
vp_Fib_R = velocity(CR,'density',rho);
vp_Fib_H = velocity(CH,'density',rho);

% vp fibre+lots of uniform ODF
mix = 2*uniformODF(CS)+cfib

[CV, CR, CH] = calcTensor(mix,E);
vp_mix_V = velocity(CV,'density',rho);
vp_mix_R = velocity(CR,'density',rho);
vp_mix_H = velocity(CH,'density',rho);

%plot things
plot(VP_SC,'antipodal','minmax')
mtexTitle('vp single')
nextAxis; nextAxis;

% plotPDF(cfib, Miller({0,0,0,1},{1,1,-2,0},CS));
% nextAxis;  

plot(vp_Fib_V,'antipodal','minmax')
mtexTitle('vp fibre Voigt')
nextAxis;
plot(vp_Fib_R,'antipodal','minmax')
mtexTitle('vp fibre Reuss')
% nextAxis;
% plot(vp_Fib_H,'antipodal','minmax')
% mtexTitle('vp cfib H')
nextAxis;

% plotPDF(mix, Miller({0,0,0,1},{1,1,-2,0},CS));
% nextAxis; 
plot(vp_mix_V,'antipodal','minmax')
mtexTitle('vp fib+uni Voigt')
nextAxis
plot(vp_mix_R,'antipodal','minmax')
mtexTitle('vp fib+uni Reuss')
% nextAxis
% plot(vp_mix_H,'antipodal','minmax')
% mtexTitle('vp mixf H')
mtexColorbar

Cheers, Rüdiger

[lfgmorales]

Hi Rüdiger

Can confirm that, plotSeismicVelocities also does not help when plotting CV,CR or CH

What sort of physics would explain that I wonder :-D

But it seems related to the mixtures of ODFs, rather then the tensor, as far As I can see

cheers

Luiz

Luiz F. G. Morales ScopeM & Structural Geology and Tectonics Group ETH Zürich

---

From: Rüdiger Kilian @.***> Sent: 22 March 2021 20:03:05 To: mtex-toolbox/mtex Cc: Subscribed Subject: [mtex-toolbox/mtex] BUG: velocities computed from stiffnessTensors from ODF with uniform components is wrong (#905)

Hi Ralf, when we have an ODF which contains mixtures with uniformComponents, the computed stiffnessTensors, respectively following results e.g. velocities can be higher or lower then single crystal values. I don't think physics allows this. Does it?

CS= crystalSymmetry('-3m', [5.323 5.323 7.272], 'X||a*', 'Y||b', 'Z||c');

rho= 2.5155; %g/cm³

N=[[229.08 89.04 13.56 0 0 0];...

[89.04 229.08 13.56 0 0 0];...

[13.56 13.56 45.84 0 0 0];...

[0 0 0 12.77 0 0];...

[0 0 0 0 12.77 0];...

[0 0 0 0 0 70.02]];

E=stiffnessTensor(N,CS)

% vp single crystal

VP_SC = velocity(E,'density',rho);

% vp fibre

cfib = fibreODF(Miller(0,0,0,1,CS),yvector,'halfwidth',30*degree)

[CV, CR, CH] = calcTensor(cfib,E);

vp_Fib_V = velocity(CV,'density',rho);

vp_Fib_R = velocity(CR,'density',rho);

vp_Fib_H = velocity(CH,'density',rho);

% vp fibre+lots of uniform ODF

mix = 2*uniformODF(CS)+cfib

[CV, CR, CH] = calcTensor(mix,E);

vp_mix_V = velocity(CV,'density',rho);

vp_mix_R = velocity(CR,'density',rho);

vp_mix_H = velocity(CH,'density',rho);

%plot things

plot(VP_SC,'antipodal','minmax')

mtexTitle('vp single')

nextAxis; nextAxis;

% plotPDF(cfib, Miller({0,0,0,1},{1,1,-2,0},CS));

% nextAxis;

plot(vp_Fib_V,'antipodal','minmax')

mtexTitle('vp fibre Voigt')

nextAxis;

plot(vp_Fib_R,'antipodal','minmax')

mtexTitle('vp fibre Reuss')

% nextAxis;

% plot(vp_Fib_H,'antipodal','minmax')

% mtexTitle('vp cfib H')

nextAxis;

% plotPDF(mix, Miller({0,0,0,1},{1,1,-2,0},CS));

% nextAxis;

plot(vp_mix_V,'antipodal','minmax')

mtexTitle('vp fib+uni Voigt')

nextAxis

plot(vp_mix_R,'antipodal','minmax')

mtexTitle('vp fib+uni Reuss')

% nextAxis

% plot(vp_mix_H,'antipodal','minmax')

% mtexTitle('vp mixf H')

mtexColorbar

[seismo_not_right]https://user-images.githubusercontent.com/11697557/112043932-7e229080-8b49-11eb-979b-04d832aea6a3.png

Cheers, Rüdiger

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHubhttps://github.com/mtex-toolbox/mtex/issues/905, or unsubscribehttps://github.com/notifications/unsubscribe-auth/ALE5IN5UMK32WZIAPCRF7W3TE6H6TANCNFSM4ZTWROCA.

[ { @.": "http://schema.org", @.": "EmailMessage", "potentialAction": { @.": "ViewAction", "target": "https://github.com/mtex-toolbox/mtex/issues/905", "url": "https://github.com/mtex-toolbox/mtex/issues/905", "name": "View Issue" }, "description": "View this Issue on GitHub", "publisher": { @.": "Organization", "name": "GitHub", "url": "https://github.com" } } ]

[ralfHielscher]

Again, I think this is due to missing normalization.

[kilir]

Oh yes, thanks, that true and the point. Glad that physics persist!