I used a commercial software LS Dyna in my PhD research. My mechanical engineering PhD research was completely modeling based, no experimental. I have three journal papers, which are in surface engineering journals and not computational modeling journals.

I don't know all details about the software/numerical method I am using. I am not comfortable in teaching or explaining someone how the software actually does the calculations. I always used the software as a tool and thought that a superficial knowledge about the software would be adequate. That's the reason I focused more on research rather than taking courses to fill the knowledge gaps.

Now, I am about to defend and I am feeling that I will fail as the committee might ask me mathematical questions and I will not know how to answer. I am screwed.

I am doing my PhD in top 30 universities in the world. I did not have to take coursework during my PhD. I don't know what will I do if that's the reason I fail my PhD defense. I am freaking out. Sorry for the rant. How should I tackle this problem?

  • 1
    You should have had this discussion about understanding, or not, the numerical methods with your advisor long ago...
    – Solar Mike
    Nov 25, 2019 at 9:47
  • @SolarMike I had. She said it's fine to not know the inner workings of the software completely as ours is an applied field and we are using it as a tool. She said that, to do research one only needs the ability to use it correctly and mathematical development capabilities is not necessary. She said that can be developed later She also said that I have done a good job in my PhD. Now I regret taking her advice.
    – Jaggy
    Nov 25, 2019 at 9:50
  • @Jaggy : You have delivered 3 Journal papers. Now Your advisor will make sure you get the degree. Don't think too much and prepare only what you have done in research and not what you don't know.
    – Rajesh D
    Nov 25, 2019 at 12:00
  • That down vote was mine. this is not really a question.
    – Rajesh D
    Nov 25, 2019 at 12:02
  • @RajeshDachiraju Thank you for your comment. I understand the downvote.
    – Jaggy
    Nov 25, 2019 at 12:03

1 Answer 1


Your advisor is giving you good advice. She presumably knows what is required here and will come to your aid if any issue arises. An advisor supporting a candidate in a defense happens in some situations. And the advisor probably knows more about all of your work than any other committee member.

But if you are using a standard tool in a standard way then it is unlikely that the tool has affected your results. Hopefully you can refer to other, published, studies that use the tool or even marketing material that point to its verification.

Your field isn't the development of such tools, but their use. Had it been tool development for ME then you'd have a problem. Your undergraduate education probably gives you a general basis for the underlying math/physics/dynamics/whatever.


Note that my experience is in situations in which the advisor is either a committee member or sits as an advocate. I recognize that there are other traditions in which the committee is independent. But even then, you are unlikely to be strongly challenged for not having details of the tool at hand, beyond general ME principles of analysis.

  • I hope that I pass my defense. In my university, the committee is independent but the advisor has a strong role to play with the outcome of a defense. I am curious, what would be best way to say I don't know something incase a committee member asks me a difficult question?
    – Jaggy
    Nov 25, 2019 at 12:38
  • Honesty usually works best. Trying to fake it is usually obvious. But falling back on principles is probably enough. But, you can say that the tool, for example is "standard" stuff. Like anything else, a defense need not be perfect in every detail to be successful.
    – Buffy
    Nov 25, 2019 at 12:52
  • +1 being able to refer to studies that show this tool is reliable Nov 26, 2019 at 2:41

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .