# Beginners Shell validation question **URL:** \ **Category:** Uncategorized\ **Created:** [February 29, 2024, 12:33pm UTC](https://calculix.discourse.group/t/beginners-shell-validation-question/2120 "2024-02-29T12:33:18Z")\ **Posts on this page:** 6\ **Page:** 1
**Author:** ![chep](https://yyz2.discourse-cdn.com/free1/user_avatar/calculix.discourse.group/chep/32/2725_2.png) [@chep](https://calculix.discourse.group/u/chep)\ **Post date:** [February 29, 2024, 12:33pm UTC](https://calculix.discourse.group/t/beginners-shell-validation-question/2120/1 "2024-02-29T12:33:18Z")
Good day all, I have been exploring the capabilities of ccx/cgx and I have a very basic question on the use of ccx and cgx with respect to shell elements. I have tried running a few of the test validation cases with v 21, and on my system the cgx output looks to have large displacement errors. I am running on a windows box within a MINGW64 Bash shell using the Calculix running environment. As an example, if I run ccx on shell1lin.inp from the Calculix2.21 test set > > ccx shell1lin Then to view the results > > cgx shell1lin.frd I see this: ![shell1lin](https://global.discourse-cdn.com/free1/uploads/calculix/original/2X/1/1f0eb751103c8aa0dc9061616f00d6a57f463d0d.jpeg) It is as if the internally generated 3D solid nodes for the shell are going off into space… I have tried a few shell input models, and all have similar distorted views. I am obviously doing something very wrong. Any help would be greatly appreciated. Thanks, Chris ---
**Author:** ![Calc\_em](https://avatars.discourse-cdn.com/v4/letter/c/43a26b/32.png) [@Calc\_em](https://calculix.discourse.group/u/Calc_em)\ **Post date:** [February 29, 2024, 12:52pm UTC](https://calculix.discourse.group/t/beginners-shell-validation-question/2120/2 "2024-02-29T12:52:21Z")
Well, I see this: ![shell1lin](https://global.discourse-cdn.com/free1/uploads/calculix/original/2X/6/6325131e9bf8712a83b3e912bd7c372e938acd29.png) It seems to be an issue with GraphiX. ---
**Author:** ![chep](https://yyz2.discourse-cdn.com/free1/user_avatar/calculix.discourse.group/chep/32/2725_2.png) [@chep](https://calculix.discourse.group/u/chep)\ **Post date:** [February 29, 2024, 3:49pm UTC](https://calculix.discourse.group/t/beginners-shell-validation-question/2120/3 "2024-02-29T15:49:18Z")
Interesting. So, it looks like your cgx is functioning OK? I tried using one of the U3S tests as well. And I get this using : ccx lin\_stat\_press\_load then cgx lin\_stat\_pres\_load.frd ![linstat](https://global.discourse-cdn.com/free1/uploads/calculix/original/2X/0/0d8743173701a994ec9d75d469da6214cfb25f25.jpeg) I am unclear on what to try next. ---
**Author:** ![chep](https://yyz2.discourse-cdn.com/free1/user_avatar/calculix.discourse.group/chep/32/2725_2.png) [@chep](https://calculix.discourse.group/u/chep)\ **Post date:** [February 29, 2024, 5:06pm UTC](https://calculix.discourse.group/t/beginners-shell-validation-question/2120/4 "2024-02-29T17:06:55Z")
Hi again, Well, after digging some more I located this thread by Calc\_em[https://calculix.discourse.group/t/help-on-installing-calculix-2-19-on-windows/1125/3](https://calculix.discourse.group/t/help-on-installing-calculix-2-19-on-windows/1125/3) This version seems to work quite well. I tried one of the more complex models I am working with now, which contains S4, S3 Mass and T3D2 elements in it. With a distributed pressure load, I get these results, which look quite good. ![rv8](https://global.discourse-cdn.com/free1/uploads/calculix/original/2X/1/1d409bcbf3343af251903c9bbca42eb0fb3a5216.jpeg) I see this v2.19 builds are quite different from the 2.21 builds… I will have to dig a bit more now, I do not like mystery fixes… ---
**Author:** ![vicmw](https://avatars.discourse-cdn.com/v4/letter/v/6de8d8/32.png) [@vicmw](https://calculix.discourse.group/u/vicmw)\ **Post date:** [February 29, 2024, 11:38pm UTC](https://calculix.discourse.group/t/beginners-shell-validation-question/2120/5 "2024-02-29T23:38:42Z")
Those results look like the 3 digit exponent problem that happens with some builds on Windows. Floating point values in the .frd file should have 2 digit exponents so if you see 3 then it’ll be that. It’s extra suspicious that your CGX is showing 3 digit exponents in the screenshot. One fix is set the environment variable PRINTF\_EXPONENT\_DIGITS = 2 ---
**Author:** ![chep](https://yyz2.discourse-cdn.com/free1/user_avatar/calculix.discourse.group/chep/32/2725_2.png) [@chep](https://calculix.discourse.group/u/chep)\ **Post date:** [February 29, 2024, 11:55pm UTC](https://calculix.discourse.group/t/beginners-shell-validation-question/2120/6 "2024-02-29T23:55:54Z")
Perfect. Thanks Victor. I had noticed that difference between the two FRD files, and was having a peek at the frd code… The windows build published on [dhondt.de](http://dhondt.de) works without issue… just my gcc build has an issue. I’ll set the switch and rebuild.