Computational Biomechanics - Science topic

Hi Haidar,

Yes, Specially in the recent version of this software they add a specific module entitled "FSI". So, right now this is the best. I totally agree with you.

Bur between Ansys CFX and Ansys Fluent, which one is better?

Best,

Seif

Hello,

I am having a similar issue. Was this resolved?

Great, thank you. I'll send you a message.

It is a tool that is still in development for the collection of anthropometric data, some studies have been carried out comparing this technology with the traditional method.

Hi,

One thing you can try is calculating the angle between the vectors that run along the helices; this can be done with cpptraj. First, read in your topology and trajectory:

parm MyParm.parm7

trajin MyTraj.nc

Then you would create the vectors:

vector Helix1 <mask1> <mask2>

vector Helix2 <mask3> <mask4>

Here, mask1 and mask2 are atom masks defining each end of the first helix and mask3 and mask4 are atom masks defining each end of the second helix. Then you can use the 'vectormath' analysis command to calculate the angle between the vectors:

vectormath vec1 Helix1 vec2 Helix2 dotangle out H1-H2-angle.dat

Hope this helps,

-Dan

Hi. you can use CST studio frequency/time domain.

Hi Marcelo. We are trying something new out in this area. You might want to take a look at http://academic.udayton.edu/murray/DIMLAB/pubs/conference/2015/T2OC-3.pdf. Unfortunately, this isn't our most recent work which streamlines the process significantly compared to the work presented in that paper. That publication is currently under preparation, but you can find most of it in an upcoming dissertation. I can supply a draft if interested.

Parameter selection is extremely important to get the solution of the problem. I like you to have a look at the following article

Thank You Abdul Aziz , Prasannah , Navid and Zdenka for your suggestions.

Based on the literature survey i found that most of the works on different bones have used Linear models for static studies and other constitutive models such as concrete damage plasticity, hyperfoam , power law models etc for dynamic studies.

My suggestation is if you have access to CATIA V5 then in its wireframe mode you can generate the surfaces and easily work on it. I have tried with it in one of my online downloaded geometry. 

Hi Konstantinos,

at first sight, your model seems to be correct. I do hope it will not be too embarrassing: Did you perform some convergence tests with "your" organs? As I mentioned before contact is really nasty. Moreover, you have complex contact surfaces.

I recommend a stepwise procedure. Perform a contact analysis rigid body with rigid body motion against the bladder, e.g., with a sphere as contactor. If you get realistic deformations, try the contact deformable sphere against the (deformable) bladder. The problem in your model is the multiple contact. You do not have very much control by mere intuition.

If you have already done such tests, please forget the previous paragraph.

Contact is lots of trial and error, mostly error.

Greetz, hp

I am not sure what kind of data you have right now. Maybe you can also check the Musculoskeletal model plus 3D motion capture to obtain the musculotendon length. We use Simulink to model this process in order to know the muscle fiber length from the already known musculotendon length.

Hi, I suggest the well documented and free software MeVisLab at www.mevislab.de/

Hi, I suggest the free software Lsdyna LS-PrePost at http://www.lstc.com/products/ls-prepost.

Hi Abd,

the following article seems to be related to your topic:

Gasser, Ogden, Holzapfel -  Hyperelastic modelling of arterial layers with distributed collagen fibre orientations. J R Soc Interface. 2006 Feb 22; 3(6): 15–35. http://dx.doi.org/10.1098%2Frsif.2005.0073

Vladimir

Hi Subhash,

1. that depends on the valve you're looking at.

120mmHg is the aortic pressure during end-systole. The aortic valve is open, thus there is no load on it (if wall shear stress or other are neglected). The main load on this valve is reached shortly after valve closure. Here is a picture were you can see some pressure gradients during the heart cycle: http://howmed.net/physiology/cardiac-cycle/  The max load on the aortic valve occurs after valve closure (pAorta - pLeftVentricle: ~100 mmHg).

Another question is if you are looking at healthy or diseased hearts.

2+3. You can find several publications on this topic I guess. Maybe start with the book 'Biomechanics - Mechanical Properties of Living Tissues' or '

Biomechanics: Motion, Flow, Stress, and Growth', both by Yuan-Cheng Fung

Dear Zdenka,

I'm devoloping a model of spinal interneurons and motoneuron network.

I need a musculoskeletal model for validating my model.

My idea was to use MSMS (i have provided the link in the previous post).

I'm using for the musculoskeletal model the parameters reported in the paper "Model-Based Sensorimotor Integration for Multi-Joint Control: Development of a Virtual Arm Model"

Each dendrimer has its own structure. we  have to draw manually in chem bio draw.

In our article:

We analized differences in angles measured in horses using 3 cameras. One of these (the middle camera) was used to analyze 2D and the other 2 for 3D reconstruction. Although there were no big differences in the amount of angle, this differences were statistically signifcant.

I recommend you to construct a structure with a known angle (for example a square - 90º - with four markers) and move it in the space where the movement to analyze will be recorded.

Regards.

From engineering point of you are interested only in the intervertebral fatigue effect. It is a viscoelastic element so I think the fatigue effect can be only predicted implementing a dynamic FE analysis with a viscoelastic hypothesis on the disc. I think it is the only way because of experimental tests could be widely affected by the cadaveric state of the spine (a large reduction of water content for example). So the first step is the recovering of viscoelastic model of disc material and the implementation of the FEA model. From the loads point of view you need of the proper model too. The correlation is the lack point for this reason I suggest to apply a preliminary analysis and test on a cadaveric intervertebral disc in order to validate the FEA model and then perform the FEA using the actual model of intervertebral disc and external load (previous perform an adjusting of the cadaveric model if needed). Here you can find some information: Article: Comparison of structural behaviour of natural inter-vertebral disc and a phema-pcl-pet system using a finite element analysisM. Bellucci, L. Di Palma, L. Ambrosio, A. Apicella

International SAMPE Symposium and Exhibition 11/2000.

Regards

Dear Mr. Dhingra,

This highly depends on your target parameters and the accuracy that you want. I myself use an inverse dynamics approachs by means of a comemercial software package (AnyBody - AnyGait). They have devloped some models of gait and they are quite robust. If you refer to gait models in terms of Motion Capture Marker Protocolls, I myself use an extended PlugIn - Gait protocoll. If you are (like me) interested in the biomechanics of gait, you will need at least the markers on every segment (foot, tibia, femur, pelvis). Only that way you can completely define movement in all three dimensions, without having to constrain the joints (ankle, knee, hip) any further than anatomy dictates. Attch the markers to bony landmarks on the body that you can find easily, that way you ensure repeatability amongst subjects.

Talking about computational efficiancy you want to use an inverse dynamics approach. This yields comparable results to forward dynamics, while being much faster. see:

1. F.C. Anderson and M.G. Pandy: “Static and dynamic optimization solutions for gait are practically equivalent.” J. Biomech. vol. 34, no. 2, pp. 153–61, 2001.

Hope that helped,

all the best,

Tim

Hi Harsha,

I am really convinced that "composite-based" modeling approaches are much more effective to describe the non linear constitutive behavior of soft tissue, because they allow to establish constitutive responses that are not dependent on phenomenologically driven choices but that straight derive from the response of the tissue constituents at different scales and from their arrangement. Accordingly, model parameters in this case can be strictly related to quantities experimentally  measurable and with a clear biophysical meaning, opening towards realistic patient-specific models. Nevertheless, as we proved in many recent papers (you can find them in my profile), such a "structural" multiscale approach has to be combined with a proper modelling of collagen nonlinearities from nano (moecular scale) to microscale (fiber scale).

Regards

GVairo

First, you should make sure that your routine UINTER is actually called by Abaqus - just add a

write(*,*) 'called now'

somewhere to the routine (the output should go either to the dat or msg-file and also to the screen if you run interactively). You can also use this to find out what values of rdisp are passed if the routine is called.

In addition, changing the lopenclose-variable should trigger a severe discontinuity iteration. Does this actually happen? (look at the sta-file)

Alternatively, try to create a model with only a very small number of elements or try to use your UINTER together with one of the example files from Abaqus in the verification manual.

Finally, in Fortran you should not write

-2.65258*10**6

but

-2.65258*E6

Hi jacobus

sorry for the delay of my answer.

I never compare plugin gait and opensim model. I should do it.

You have to know that kinematic calculations of these models are different.

opensim considers a-priori definite joint axes and uses an inverse kinematic optimization (Lu, T., & O’connor, J. (1999). Bone position estimation from skin marker co-ordinates using global optimisation with joint constraints. Journal of Biomechanics, 32, 129–134.)

Plugin gait only considers ball joint at all articulation and extracts cardanic angles.

No model can be a golden approach !

regards

Fabien

You could use a tool like OpenSim to find the centers of mass in the posture you are looking for and then add up the moments of inertia using Steiner's law.

I suggest you contact vito.starc@mf.uni-lj.si if he has it or he knows someone who does...Good luck! Dusan

I have some experience with BTS IR system and they do have a software called BTS Clinic, extremely easy to use for begginers... You should introduce a complete protocol but once you have it done, it can be used for non-experienced biomechanical professionals... Maybe you can have a look on it also... Vicon and Qualysis are both excellent options for biomechanics!

Although the theoretical equation between an angular velocity w and a rotation matrix R is given by S(w)=dot_R*R^T so that dot_R=S(w)*R, we should always try to avoid applying the first-order approximation directly on the time-derivative of a matrix, such as using [R(i+1)-R(i)]/dt=S(w)*R(i) to update the orientation R(i). A better way in this particular problem is to apply the first-order approximation on a scalar, such as the rotation angle phi, instead of the rotation matrix R, as what I procedurized in my first answer by taking advantage of the angular velocity definition w=dot_phi*k with a temporarily fixed unity axis k and letting dot_phi=dphi/dt for the scalar angle phi. Then, utilizing the Rodrigues formula to find the corresponding rotation matrix R(dt), as given in Step (2) of my procedure. Finally, through the multiplication R(i)*R(dt)=R(i+1), the orientation is updated. Using this way, we can not only achieve an acceptable numerical solution, but also keep the orthogonality invariant.

In general, as we all experienced, to implement a derivative in computer, it needs a past-value memory in order to achieve a higher-order numerical solution. In contrast, to implement an integration, we can easily get a fourth-order approximation, such as using the well-known Runge-Kutta algorithm. Therefore, it has to be more careful to manipulate the numerical derivatives in computer programming.

you might find some useful citations in the following paper https://www.researchgate.net/publication/236877584_Third-Order_Muscle_Models_The_Role_of_Oscillatory_Behavior_in_Force_Control?fulltextDialog=true&ev=prf_pub_xaddfulltext

Hi,

it may seem weird, but I think you could search for software in the WWW that is normally used to reconstruct documents that have been torn into pieces. This software should be able to solve the puzzle by first detecting the "fracture" pattern. I have no idea if this works out, but maybe it is worth a try ...

Good luck

Uli H.

Rıza,

David Winter made this issue clear in his pioneering book (Biomechanics and Motor Control of Human Movement, 2nd Edition, page 78-79).

“Considerable confusion exists regarding the relationship between joint reaction forces and joint bone-to-bone forces. The latter forces are the actual forces seen across the articulating surfaces and include the effect of muscle activity. The bone-to-bone force equals the active compression force due to muscle plus joint reaction forces.”

In your case, hip contact force equals to bone-to-bone forces. See the attached file for further details.