[javert]

Hi everyone. I'm working on some chemical engineering problems but after some unnecessary struggle I have decided to acquire a software at last and stop having to do handwriting to tinker with equations and having to recur to Excel to perform heuristics like Euler or Runge-Kutta methods. There seem to be a handful of Computer Algebra Systems out there and I don't have the time or the money to try all of them, so I'm asking here to people who have experience with them which one would you recommend to fellow chemical engineers. It doesn't have to be free, but it does need to be good (and in the times of "Free to play", they are almost opposite).

To put on context, right now I'm dealing with calculations regarding a continuous bioreactor where microbial growth is modeled with Monod equations and the cell yield Y is constant (pretty basic model I guess, yet it already involves 4 equations with some of those being quadratic for some variables). Ideally, I want the following from the software:

• I would like a software that is able to recognize variables, equations and systems of equations as proper objects, so I can ask the software to solve a equation or system of equations for some of them when the others take fixed values.
  
• A graphic user interface for editing those equations and for assigning the fixed values to to variables in order to estimate others. I want to skip the use of command lines as much as possible (for example, having to manually set Volume=2 m3, Flow=1.5, etc) becaus that also involve a lot of hours of consulting on manuals the proper syntax of functions and their inputs. I have used R once for statistic purposes and I do not wish to experience that pain again.
• A built in unit converter (for example, the variable object could include a unit), to avoid having to do those converions outside the CAS.
• As optional, 3D vectorial and tensor calculus included. I might be interested on doing some basic heat or mass flow in the future.

[Enthalpy]

Hi javert,

I don't know if there is such a software specific to chemistry, but I use more general ones for algebra:

• Mathcad computes numerical solutions and makes also user-guided symbolic transformations:
• Maple makes autonomously symbolic transformations and computes also numerical solutions.
• Many more exist, the free ones being mainly for Linux. These may even integrate with Latex to make cute reports.

That is, to the integral of cos(x), Mathcad answers 0.123456 and Maple sin(x) - but Maple can evaluate numerical values and Mathcad can do symbolic computations if given more directives from the user.

Graphs and sets of linear equations are nothing special for both of them, which are brutally more powerful than that. I didn't re-check if they know units natively, but if not, libraries exist. Both are extensible through user programming too but work well without. Both are decently easy to learn.

I have the versions 2000 pro of Mathcad (last version without the infamous C-Dilla) and 7 of Maple; Maple 9 seems to need no activation but 9.5 for sure needs one. Mathcad 2000 installs to Win Xp and possible more (and from Win95a for sure), Maple 7 to 64b-Seven and maybe more (and from Win95a for sure). Older versions are perfectly useable (I had one on three diskettes).

I bought them used through eBay, they are not so common and are language-specific. Depending on your location, you have to explain to the seller that the Cocom is outdated (...little after the Soviet union disappeared) and US law allows now to export such software.

[Irlanur]

Numerical: MATLAB (does in principle also allow symbolic calculations, but not so straight-forward I think)
symbolic: Mathematica

at my uni, all the chemical engineers extensively use MATLAB in the curriculum as well as in research. In magnetic resonance calculations and simulations, MATLAB and Mathematica are also used a lot.