2 words of Caution in how to use GRT

[dstoenner]

First) A true story. My friend got his 243 rebarreled (by me) and he took a couple of loads to the range. Unfortunately he hit high pressure with one but not terrible. He gave me his velocity and powder level and I plugged them into GRT. I then did an OBT and yes GRT said there were 103%. I then ran an OBT that would be lower pressure and gave that to my friend. He loaded some at the projected target value and some at .2 gn higher. When he shot them they didn't group and when he gave me the velocities, they were 200 fps slower than predicted.

The light bulb came on so I inputted the new velocities and did an OBT of the new values and the powder parameters Ba and K were lower values and predicted the new powder points perfectly and had a Pmax -15% node at 2 gn 1 gn more.

SO here is the caution. My conclusion is that OBT is a piecewise linear approximation to the actual curves of BA and K which are not linear. And where you run OBT WRT the center of the curve will give you radically different values as you try to extend it lower or higher.

My recommendation for a strategy here to find a good first order solution is to take the base value of powder from the table with att of you inputs on your gun and case volume and find a node somewhere in the 15 to 25% band. Now load 5 rounds at that value and 5 rounds at 1 gn more or less to establish to points on the curves and run OBT a second time to get a better understanding where your nodes might be. Then test the node with a ladder that has that node somewhere in the steps. You should get a good node to then start with.

Second issue. I have also figured out that GRT has some filter on when it will modify the powder parameters curve fit and when it will just leave them alone. The only way to force it to do a real curve fit to you parameters is to go back and reselect the powder from the internal tables and then do the OBT. I back annotate all my runs If you do not back annotate the Ba and K values then you don't have to worry it will ALWAYs curve fit for that run of OBT.

Hope this helps

David

 

[divingin]

It's software, and it's a predictive tool. Simulations are just that. They approximate what the given parameters *may* yield.

It's pretty foolish to think that you can run whatever it says is safe (that actually holds true for Quickload, and data out of a manual, as well), given that you are running different component lots, in different conditions, on different hardware, with different loading practices. Always work up to a load, especially ones near the top of the pressure curve, where a problem can get expensive quickly.

 

[Chestnut]

Second issue. I have also figured out that GRT has some filter on when it will modify the powder parameters curve fit and when it will just leave them alone. The only way to force it to do a real curve fit to you parameters is to go back and reselect the powder from the internal tables and then do the OBT. I back annotate all my runs If you do not back annotate the Ba and K values then you don't have to worry it will ALWAYs curve fit for that run of OBT.

Can you expand on what you mean by this a bit more? I'm just not sure where it fits into my steps that i've outlined below.

1. I'll select a powder charge shoot 5 shots enter the velocity into a measurements.
2. Next i make sure the powder matches that measurements and run OBT.
3. Note OBT charge
4. Enter K and Ba values
5. change charge to OBT sugessted, load 5 shots
6. shot said values recording velocity and start over at step 1.

 

[J_idaho]

After using GRT enough you realize there’s to many changing variables to hit the nodes every time. I believe your starting pressure is never right because the burn rate never extends far enough. So you know some variable is off. It is a nice tool for starting points once you kinda of know how the program is going to react. It will never be accurate until you put enough data in that you already know where to load anyways. That’s my 2 cents.

 

[dstoenner]

Can you expand on what you mean by this a bit more? I'm just not sure where it fits into my steps that i've outlined below.

1. I'll select a powder charge shoot 5 shots enter the velocity into a measurements.
2. Next i make sure the powder matches that measurements and run OBT.
3. Note OBT charge
4. Enter K and Ba values
5. change charge to OBT sugessted, load 5 shots
6. shot said values recording velocity and start over at step 1.

The problem that i have seen is after you back annotate the first time is that obt will not change them the second and third time you add data and you are getting close but incorrect data to establish what might be a node

David

 

[dstoenner]

After using GRT enough you realize there’s to many changing variables to hit the nodes every time. I believe your starting pressure is never right because the burn rate never extends far enough. So you know some variable is off. It is a nice tool for starting points once you kinda of know how the program is going to react. It will never be accurate until you put enough data in that you already know where to load anyways. That’s my 2 cents.

Yes that is true. Once you put in a ladder of 5 shots varying by .2 or .3 gns per step then the accuracy and curve fitting are very good. Until then it is a SWAG

David

 

[SeabeeKen]

First) A true story. My friend got his 243 rebarreled (by me) and he took a couple of loads to the range. Unfortunately he hit high pressure with one but not terrible. He gave me his velocity and powder level and I plugged them into GRT. I then did an OBT and yes GRT said there were 103%. I then ran an OBT that would be lower pressure and gave that to my friend. He loaded some at the projected target value and some at .2 gn higher. When he shot them they didn't group and when he gave me the velocities, they were 200 fps slower than predicted.

The light bulb came on so I inputted the new velocities and did an OBT of the new values and the powder parameters Ba and K were lower values and predicted the new powder points perfectly and had a Pmax -15% node at 2 gn 1 gn more.

SO here is the caution. My conclusion is that OBT is a piecewise linear approximation to the actual curves of BA and K which are not linear. And where you run OBT WRT the center of the curve will give you radically different values as you try to extend it lower or higher.

My recomendation for a startegy here to find a good first order solution is to take the base value of powder from the table with att of you inputs on your gun and case volume and find a node somewhere in the 15 to 25% band. Now load 5 rounds at that value and 5 rounds at 1 gn more or less to establish to points on the curves and run OBT a second time to get a better understanding where your nodes might be. Then test the node with a ladder that has that node somewhere in the steps. You should get a good node to then start with.

Second issue. I have also figured out that GRT has some filter on when it will modify the powder parameters curve fit and when it will just leave them alone. The only way to force it to do a real curve fit to you parameters is to go back and reselect the powder from the internal tables and then do the OBT. I back annotate all my runs If you do not back annotate the Ba and K values then you don't have to worry it will ALWAYs curve fit for that run of OBT.

Hope this helps

David

Gordon said that they found obt to only be correct 40% of the time. My experience with it has been better than that. GRT is a prediction, not fact

 

[Tangent]

I've not used it myself but my experience of computer systems, data and simulations is it's all dependent on the input. Many stacking variables of potentially uncertain values gives quite a wide min/max. I can see it being useful to weed out the extremes but once narrowed down I don't think there's any substitute for practical testing.

As an aside I suspect there's a lack of this in the world right now. Throw some stuff into the mix and see what happens, after all most great discoveries were accidents.

 

[Webster]

Gordon said that they found obt to only be correct 40% of the time. My experience with it has been better than that. GRT is a prediction, not fact

I am pretty good with computers but never used a loading program. I don't trust them. To many variabls and every rifle is different. If you cannot find a good load you probably have an average barrel and bad bullets. Everything depends on what you call a good group. Both my hunting rifles shoot just about any load under .400" groups. Never had any group strung out in one direction.

 

[Bat Rastard]

OP, I don't know if startegy is a word, but I will be using it going forward.

 

[dstoenner]

OP, I don't know if startegy is a word, but I will be using it going forward.

Opps. Corrected and recommendation was spelled wrong. I never was good spelling but math is my language.

David

 

[Orr89rocz]

Ive learned its only a simulation and alot of assumptions need made in order to use it that might not hold consistent thru a charge ladder, and it makes sense i think as i guess burn rate is variable with pressure. Or a combination of starting pressure along with the sebert factor may change? Lower end charges may not show complete burn and require a different burn rate, but as i match that performance and increase charge, my higher end charge weights dont align with velocity there, so again play with some of the factors and burn rate etc. im still new to GRT and have been using quickload. I found both to be helpful but for some cartridges and powders to be very inaccurate using base assumptions and for others somewhat closer to what i seen.
When dealing with data that doesnt exist in books, powder/bullet weight/cartridge combos that arent used much, you gotta take it with a grain of salt i guess. Its funny cuz online youtube experts seem to use it religiously as gospel to compare powders and cartridges with out fully testing it and draw conclusions from it. Gotta be careful with that imo

 

[Fclassfool]

I don't know how I managed to reload for 35 years and not need some computer program to tell me how....