Dear sir,

I think you are talking about mesh sensitivity, infact it's nothing but mesh convergence study. In this, you need to analyse concrete models for different mesh density and check compressive strength for every mesh pattern till the solution converged. Thanks. Hope it will be beneficial for you.

Vary one parameter at a time and observe the response i.e. univariate analysis

If you have a deterministic system, there are multiple established methods. It may require some programming, but nothing too difficult.

You may change a little parameters of your system, or parameters of external excitation. If behavior of the system was changed a little too, the system is structurally stable. If not, the system is structurally unstable, or you have some resonance situation.

Sensitivity analysis is a technique that analyses which parameter is more effective for results. There are different type of methods used for sensitivity analysis:

1. OAT( one at a time) techniques

2. Monte Carlo techniques

3. Differential sensitivity analysis

4. Regression analysis

5. Variance-based sensitivity analysis

6. Scatter plots

7. Relative-deviation method

8. Smirnov Test and so on…

solved your system numerically with your data, then varying one parameter each time.

I suggest to see the below link

https://www.youtube.com/watch?v=3wNxZcvRdPI

Regards

Sensitivity analysis can mean rather a lot of things. The first encounter I had - probably in 1984, in a course on linear optimisation - was that of the computation and analysis of shadow prices - basically analysing the potential extra gains if we are allowed more resources. In some cases those shadow prices are actually directional derivatives of a special parameterised function.

One major thing you need to know is that sensitivity analysis is a post-optimality operation, i.e., an optimal solution to a problem/model must be obtained before studying the effect of discrete parameter changes on the optimal solution - sensitivity analysis. Interestingly, there are different approaches to this as mentioned by @Ram Chandra, you may determine the best by trying.

The SUNDIALS ODE solver suite

https://computing.llnl.gov/projects/sundials

includes CVODES which solves ODE systems and includes sensitivity analysis capabilities.

In Civil Engineering field (particularly in concrete technology) it is not enough to compute individual solutions of equations created while solving a mathematical model. It should be noticed that all possible parameters should be included with their correct importance factor. This is particularly true in civil engineering applications where parameters of the model which has to study are often only imprecisely known. Hence the researcher has a task to identify those dependent critical parameters and emphasize them with correct importance factor.

Hi? You might want to have a look at one of my publications - 10.1016/j.envsoft.2020.104800

I recently conducted a similar study trying to analyze the impact/influence of a parameter in the model output! Instead of the fire simulations, you can define the equations easily in SALib python library and run the analysis.

Cheers!