Hi Rober,
Your mass density seems to be off. If you are using SI units, as it appears from Young's moduls, mass density must be on the order of 1E3. Take a look at this revised mass density and frequency range:
structuralModel = createpde('structural','modal-planestrain');
% circles
C1 = [1,0.5,-0.25,0.25]';
C1 = [C1;zeros(10 - length(C1),1)];
C2 = [1,0.5,-0.25,0.20]';
C2 = [C2;zeros(10 - length(C2),1)];
C3 = [1,0.5,-0.25,0.18]';
C3 = [C3;zeros(10 - length(C3),1)];
% geometry from edges
gm = [C1,C2,C3];
sf = '(C1-C2)+(C2-C3)';
ns = char('C1','C2','C3');
ns = ns';
g = decsg(gm,sf,ns);
geometryFromEdges(structuralModel,g);
% subplot(3,1,1)
pdegplot(structuralModel,'Facelabels','on','EdgeLabels','on') % ,'EdgeLabels','on',
axis equal
% material properties
structuralProperties(structuralModel,'Face',1,'YoungsModulus',200E9,'PoissonsRatio',0.3,'MassDensity',2.7E3);
structuralProperties(structuralModel,'Face',2,'YoungsModulus',110E9,'PoissonsRatio',0.28,'MassDensity',8.7E2);
% mesh
const_h_max = 0.005;
mesh = generateMesh(structuralModel,'Hmax',const_h_max);
subplot(3,1,2)
pdeplot(mesh)
% solving
eigen = solve(structuralModel,'FrequencyRange',[-Inf,10000]*2*pi);
