T-FLEX – First Fiddle

I had mentioned T-FLEX Parametric CAD earlier and a reader wanted to know when I would review the software. I write software, not reviews, for a living. So as I fiddle around with T-FLEX I shall post stuff about the software that interests me.

On a recent business trip to Mumbai (“Mumbai Nightmares“) I met the Indian distributor of T-FLEX. He gave me the not-yet-released version 11 of T-FLEX. We didn’t have the time for a full fledged demo, which was a good thing, because I am the kind of person who likes to learn software on my own. I know its the hard way out, but I usually get a better feeling of how intuitive a product is and how easily I can get access to help when required.

T-FLEX is pretty much like any other parametric solid modeler but does a few things differently. I get the feeling that the T-FLEX user interface has been designed keeping AutoCAD users in mind. For example, to create a revolve feature, create the profile as you would in a 2D environment. Then before drawing the axis set the line type to “Center” and hit “Revolve”. Bang! You have your revolve feature. This thing actually has an AutoCAD style toolbar with layers, line types, colors, etc. in the part environment itself. The figure below shows a closed profile and the center line. The dashed construction can be easily and quickly created before you start the sketch, thanks to a very intuitive set of tools to create them.

T-FLEX is built over the Parasolid 19 modeling kernel, the latest version. So the actual solid modeling should be the best there is. I picked this image from their help file which shows that the Y-blend function supports the blend creation although a small face is in the way.

I am sure there are a lot of nice things in T-FLEX, but I doubt anything beats this next feature. They call it “Parametrization”. If you thought that parametrics makes a model intelligent, then parametrization is going to make it a genius. As in normal parametric modeling, as you peform operations on your model they are recorded in the history tree. These parameters have values. A fillet will have a radius, a hole will have a diameter, etc. In T-FLEX you can store these parameters as variables, say “R” for a fillet radius and “D” for a hole diameter. What’s more, these variables can be driven by mathematical expressions and can even be dependent on other parameters and variables. For example, fillet radius “R” can be set to twice hole diameter “D”. Finally, you can invoke something known as a Variable Editor (basically a dialog box which lists all the variables in the model), change their values and the model will be updated.

So using parametrization, you can create a variable driven model which can be used as a base for other variants of the model. You can even use this technique to automatically optimize your model. I haven’t reached there yet. So I will leave that for “T-FLEX – Second Fiddle”.

So can all this whiz bang stuff handle large assemblies? That’s precisely what I asked the distributor. He replied that a Russian aircract engine manufacturer had purchased 98 licenses of T-FLEX and was using it to design their aircraft engines.

Looks like I will be fiddling around with this piece of work for a while.

  • Scott

    I looked at T-flex a long time ago. The marketing aspect they approached at the time was to not usurp Solidworks or SolidEdge (Inventor wasn’t around yet), but to find a niche in those who make a lot of last minute, non-intuitive, modeling changes (scoped beyond design intent) but still have a parametric tool.

    It’s hard to compare what T-flex was to what is now. For example, T-flex allowed creating you entire assembly in one part, and then breaking the parts up into individual files later. That is now comfortably known as multi-part bodies using a master model approach. When they first did it, it was brand new. Their problem, no headlines about it. Their latest variable table enhancement – SolidEdge has been doing that since its conception.

    I think if T-Flex was the marketing machine in its infancy like Solidworks was/is, the market would be entirely different.

  • Scott

    I looked at T-flex a long time ago. The marketing aspect they approached at the time was to not usurp Solidworks or SolidEdge (Inventor wasn’t around yet), but to find a niche in those who make a lot of last minute, non-intuitive, modeling changes (scoped beyond design intent) but still have a parametric tool.It’s hard to compare what T-flex was to what is now. For example, T-flex allowed creating you entire assembly in one part, and then breaking the parts up into individual files later. That is now comfortably known as multi-part bodies using a master model approach. When they first did it, it was brand new. Their problem, no headlines about it. Their latest variable table enhancement – SolidEdge has been doing that since its conception.I think if T-Flex was the marketing machine in its infancy like Solidworks was/is, the market would be entirely different.

  • Deelip Menezes

    Yes, their main problem appears to be marketing. They seem to already have a good product.

    Do they even have a PR firm working for them? And if they do, what the hell have they been upto?

  • Deelip Menezes

    Yes, their main problem appears to be marketing. They seem to already have a good product.Do they even have a PR firm working for them? And if they do, what the hell have they been upto?

  • Anonymous

    I think that their “Parametrisation” exists in Pro/Engineer for a while (as expressions and family tables)…

  • Anonymous

    I think that their “Parametrisation” exists in Pro/Engineer for a while (as expressions and family tables)…