I've finally hit a GRT snag...

[Mr_Flintstone]

I've been using GRT for quite some time, and I've always had good luck with getting pressures close to what is printed in manuals when testing known loads. Once I've got my numbers working with verified loads, I then proceed to testing unpublished data. Now to my problem.

I wanted to experiment with my .32 H&R magnum, so I did the same things I always do in testing data from manuals to see how they compare to computer predictions of velocity and pressure. A problem is that GRT measures in PSI, and nearly all 32 H&R data is in CUP. To overcome that, I just used max loads that were very close to 21,000 CUP to see what their predictions were in PSI. The first powder I tried was Titegroup which was listed with an 85 grain XTP as having a pressure of 20,900 CUP, so I thought that should be in excess of 21,000 PSI, and it was 21,025 psi. That looked OK, so then I start testing other powders, and this is where things go sideways. The one I tested next was 231. It showed a published pressure of 20,800 CUP, so I thought it should give something close to Titegroup, but it didn't. It gave 16,700 PSI. Everything checked out OK, but a really low pressure, so on to the next. I tested max loads of Autocomp, HS-6, Bullseye, and Unique, and they all gave pressures between 13,000-17,000 PSI when published data showed pressures over 20,000 CUP. The only other powder that came anywhere close to published data was Universal, but it was somewhat low too. I thought this was really strange because my loads for .32 S&W Long are usually pretty close to published data. At this point I don't feel confident in using GRT to make predictions with 32 H&R Magnum.

Has anyone else had any problems with this or any other particular load in GRT?

 

[jmorris]

When people come up with calculations to predict things they are never perfect and can even be completely wrong with some inputs.

Why they seem great with the components the calculations were derived from and not so great at extremities.

 

[Mr_Flintstone]

I guess you get what you pay for. It's a shame, because I really like the program, and have always had good luck with it. So much so that I gave up on wanting to purchase Quickload a while back. I think maybe I'll pony up the money and finally buy Quickload just so I'll have a second method of prediction and modeling.

 

[GeoDudeFlorida]

When people come up with calculations to predict things they are never perfect and can even be completely wrong with some inputs.

Why they seem great with the components the calculations were derived from and not so great at extremities.

So true!
Predictive analytics is more voodoo than science. But the general public, unfortunately, trusts computer models over collected samples. It’s one of the downsides of the lingual shift which has been progressing through the last several generations. Science is just a method of testing theory, not a series of facts or proofs.

 

[AJC1]

I've been using GRT for quite some time, and I've always had good luck with getting pressures close to what is printed in manuals when testing known loads. Once I've got my numbers working with verified loads, I then proceed to testing unpublished data. Now to my problem.

I wanted to experiment with my .32 H&R magnum, so I did the same things I always do in testing data from manuals to see how they compare to computer predictions of velocity and pressure. A problem is that GRT measures in PSI, and nearly all 32 H&R data is in CUP. To overcome that, I just used max loads that were very close to 21,000 CUP to see what their predictions were in PSI. The first powder I tried was Titegroup which was listed with an 85 grain XTP as having a pressure of 20,900 CUP, so I thought that should be in excess of 21,000 PSI, and it was 21,025 psi. That looked OK, so then I start testing other powders, and this is where things go sideways. The one I tested next was 231. It showed a published pressure of 20,800 CUP, so I thought it should give something close to Titegroup, but it didn't. It gave 16,700 PSI. Everything checked out OK, but a really low pressure, so on to the next. I tested max loads of Autocomp, HS-6, Bullseye, and Unique, and they all gave pressures between 13,000-17,000 PSI when published data showed pressures over 20,000 CUP. The only other powder that came anywhere close to published data was Universal, but it was somewhat low too. I thought this was really strange because my loads for .32 S&W Long are usually pretty close to published data. At this point I don't feel confident in using GRT to make predictions with 32 H&R Magnum.

Has anyone else had any problems with this or any other particular load in GRT?

One thing I've seen but haven't used or verified is the ability to enter shot string data. I don't know if it changes the outputs or not, but it may be worth 200 dollars to see if adding your test data brings the results more in line with your expectations.... the methods of collecting data also have a big impact on old vs new data. That may very well be having a drastic effect on the result.

 

[JohnMc]

According to the formula on the Wikipedia page (worth every penny, I know), 20,000 CUP is 12,400 PSI, so apart from the near parity of the numbers for the first load, I'd say GRT is doing OK. According to the Calculator Academy site 20k CUP is 12,418 PSI.

CUP ↔ PSI Calculator - Calculator Academy

Enter the total CUP into the calculator to determine the equivalent PSI.

calculator.academy

Copper units of pressure - Wikipedia

en.wikipedia.org

As said above, modeling is just that; a model based on test data. I don't know enough about how this is handled in the GRT community but I understand that it's constantly being upgraded based on communication in the forum. If you're willing to help with data, they somehow incorporate it into the next release(s).

 

[fxvr5]

QuickLOAD has the same problem. Some of its predictions are way off. There is a warning when you open QuickLOAD; it is no substitute for pressure tested data.

 

[Hooda Thunkit]

In my experience, while GRT is very good with bottleneck cartridges, some of the data and predictions for straight-wall cartridges can be wonky.

OTOH, it nailed 38 Spl +P with HS-6 perfectly.

 

[tmd16556]

GRT also notes how well calibrated each powder is for a cartridge. Some are high confidence, others are more based on general models. On odd calibers it seems best for checking published loads. I had a 222 REM workup from data in a recent manual (either Hornady 10 or Lyman 50, I don’t remember which at the moment) that had charges at the 223 level. It was decent at confirming those as too high.

 

[P Flados]

I use Quickload. I tried to benchmark it against published data for 327 and got very inconsistent results. I have also had similar issues on other straight wall cartridges.

 

[GHinNH]

QL and GRT are (extremely complicated) predictions, that's all they are, predictions. With GRT the initial model only predicts until the (fired load) chrono results are input into the program, and OBT is run.

It has been mentioned repeatedly on the GRT Discord that it is primarily designed for rifle and has limited confidence for pistol, and to be used as a guide along with manufacturers recommendations in either case. That said, of course I use it for pistol also.

Until (you) have chrono results and OBT is run, it is pretty much a waste of time to extrapolate on initial unverified modeling.

If you choose to, I'll suggest joining GRT on Discord where questions can be answered by those who know best, and are continuing it's development.

 

[AJC1]

QL and GRT are (extremely complicated) predictions, that's all they are, predictions. With GRT the initial model only predicts until the (fired load) chrono results are input into the program, and OBT is run.

It has been mentioned repeatedly on the GRT Discord that it is primarily designed for rifle and has limited confidence for pistol, and to be used as a guide along with manufacturers recommendations in either case. That said, of course I use it for pistol also.

Until (you) have chrono results and OBT is run, it is pretty much a waste of time to extrapolate on initial unverified modeling.

If you choose to, I'll suggest joining GRT on Discord where questions can be answered by those who know best, and are continuing it's development.

Is obt a different section or just a different mode. I've seen the section to add data but not used it.

 

[Mr_Flintstone]

QL and GRT are (extremely complicated) predictions, that's all they are, predictions. With GRT the initial model only predicts until the (fired load) chrono results are input into the program, and OBT is run.

It has been mentioned repeatedly on the GRT Discord that it is primarily designed for rifle and has limited confidence for pistol, and to be used as a guide along with manufacturers recommendations in either case. That said, of course I use it for pistol also.

Until (you) have chrono results and OBT is run, it is pretty much a waste of time to extrapolate on initial unverified modeling.

If you choose to, I'll suggest joining GRT on Discord where questions can be answered by those who know best, and are continuing it's development.

I tried joining the discord site, and I keep getting this…

 

[GHinNH]

Optimal Barrel Time (OBT) is a component of GRT and is located in the header bar with the other features like bullet and powder selection. Looks sort of like a stopwatch with OBT in the middle of it.

It only works if you have added a "measurements" tab to your existing model and have input chrono results from the actual load modeled.

The more round counts added to the results tab increases the accuracy. I like to use at least 20 to get a good solid average of FPS.

Then after that, when you click "OBT" a popup will appear and it should be populated with the results from your measurements tab. Then you can hit "calculate" to make the model reflect your results, then save it as a separate file. I like to differentiate the file by putting "OBT" at the very beginning of the file name.

There is also a function to calculate powder charge changes in OBT, to increase theoretical accuracy improvement but that's a long subject. It would be best to go join up at GRT on Discord to get the right answers the first time. (PS read the manual first because they'll ask if you did, it's in the "help" dropdown under documentation)

 

[GHinNH]

I tried joining the discord site, and I keep getting this…
View attachment 1183590

Yeah... IDK anything about Apple anything. I use a laptop win10 and Firefox. I wouldn't even own a phone if I didn't have to, I'd just as soon throw the thing off a bridge.

 

[JohnMc]

I tried joining the discord site, and I keep getting this…
View attachment 1183590

Is that an advertisement link in the middle? If you just click the accept button, what happens?
Sometimes, well often really, websites have adverts which are links to other websites. If your safety settings are stringent many advert websites are loaded with trackers and such, so they get blocked, while the rest of the page displays OK. They end up looking something like that, though I can't say specifically for your instance.

 

[Mr_Flintstone]

Is that an advertisement link in the middle? If you just click the accept button, what happens?
Sometimes, well often really, websites have adverts which are links to other websites. If your safety settings are stringent many advert websites are loaded with trackers and such, so they get blocked, while the rest of the page displays OK. They end up looking something like that, though I can't say specifically for your instance.

No, that’s just the screen I get when I click to join the discord from www.grtools.de . When I click the accept button I get the message I posted above.

 

[Hooda Thunkit]

No, that’s just the screen I get when I click to join the discord from www.grtools.de. When I click the accept button I get the message I posted above.

Try a different browser. Opera, Brave, or Firefox.

Sometimes one or the other of those act up on a certain website, I'll just swap to another. Gotta admit, I've never used Safari.

And I'll go back to writing letters before using an MS browser.

 

[JohnMc]

No, that’s just the screen I get when I click to join the discord from www.grtools.de . When I click the accept button I get the message I posted above.

Try what Hooda Thunkit said. (edited to fix autocorrect)

 

[Blue68f100]

The link works with firefox browser, MBP running v12.7.1 OSX.

 

[Mr_Flintstone]

I finally got the link to work with Chrome. I’ll try to set it up tomorrow.