There have been a few threads on here about GRT. I have been using it for over 2 years now. As an electrical engineer, I am used to simulating systems to see the response. GRT is just exactly that but the "trick" is how to coax it into giving you a good answer. That took me some time to figure out and put it to practice. The bottom line is back annotation. I do not own QuickLoad basically because I am cheap but have learned how to make GRT give me good answers. That is why I am posting this that it might help others.
You can get GRT for free at grtools.de. You do have to create a user profile in order to download and it only runs on Windows or Linux
GRT requires that you describe the case, the bullet and the loaded round parameters (powder type, amount of powder and the COAL of the finished round). The problem is that when you select a powder in GRT you get a set of parameters that has been established by Gordon and his team. For me when I put in my known good round, the velocity never came to the point I was measuring.
So the trick is to get GRT to create a set of parameters that do give you a correct value. When GRT is doing this iterative curve fitting process it is taking you rifle(chamber, barrel, bore, bullet, case, powder, primer) as a whole system. The option for OBT is where this iterative process is run. If you compare the original screen to the OBT screen you will see 2 parametrs are different. Those are:
1) Isentropic exponent (k)
2) Combustion coefficent (Ba)
Getting these right then gives good answers to where the nodes are. So here is the way I have found to give the shortest and most efficent process for a good load. These is where you can do it in 50 bullets.
1) Build your system. Get as much into the given sections as possible. Fired case water volume is important.
2) Run an OBT and look for a node that is at at about 85% of you case pressure.
3) Use that node you just creaated to load 8 bullets to the parameters that are in GRT. this will be the first iteration of determining those 2 parameters. Fire all 8 taking Velocity on them. I use the first 3 to foul the barrel and then the last 5 for velocity data.
4) Open your original base file and add to the "Measurement" tab the 5 velocity values you obtained. Now in the base file set up the powder steps so that the value you used for powder in thewse 8 bullets is created. Then select that powder level in the "Results" window. This then sets the system to do an iterative solution with one known point. One is better than 0 when we started but is not good enough to get an accurate projection of a node.
So run OBT now at this powder level. Figure out which node you want to aim for and then have OBT calculate that node. As a final step to document the optimized results hit the "create a OBT file". Once that file is created then cut and paste these to 2 values back into the base file (iteration 1 is now done).
5) For iteration 2 we will now build 20 bullets. Here I take the the node predicted in 4) as the centerpoint of 5 test powder values .2gn apart. I load 6 for the lowest value and 3 for each value above that. The first 3 are to foul the barrel. Then I take and shot these on a target that has 6 bulls so I can see the placement of each set relative to each other. Taking velocity for each shot fired.
6) now we go back to the base file and put these all in the measurement tab as before. I typically over write the first group values unless that value was within the new set of 5. Now I adjust the pwder spread to have a value that was the center value of the 5 sets and select that value Now run OBT again for probably the same node we used in 4). Create and OBT file. Back annotate the 2 parameters for iteration 2 into the base file.
7) At this point I am usually done. Usually one value on the 5 targets is close to iteration 2's predicted node, the SD will be low. Some times it is between 2 of the steps and if we are really blessed those 2 values have low SD's and you can split your new derived node.
8) Now we have come to the "proof of the pudding" I now load up 20 at this new value and go shot 4-5 shot groups. If the average is what I am looking for, I am done.
At this point if I haven't come up with a good load but the sd is still low, I then try seating depth to tune it in. So far I haven't had to try that.
So we used 8 + 15 + 20 = 43 bullets and 3 range trips to get a working load with an unknown system to GRT.
Once I had the powder values ( RL-16 and Varget) for my system I now plugged in those values to a different bullet and used that as step 4 results to do a spread of 15 bullets. And yes when I run OBT and get the values again they are slightly different but much closer than step 1.
Hope this makes sense. Unfortunately my main computer is a MAC and I have a Windows computer in the garage where I run GRT but getting screen shots into this MAC is cumbersome. But I think that if you are a novice to GRT once you have gotten through steps 1 and 2, the rest will flow easily.
David
You can get GRT for free at grtools.de. You do have to create a user profile in order to download and it only runs on Windows or Linux
GRT requires that you describe the case, the bullet and the loaded round parameters (powder type, amount of powder and the COAL of the finished round). The problem is that when you select a powder in GRT you get a set of parameters that has been established by Gordon and his team. For me when I put in my known good round, the velocity never came to the point I was measuring.
So the trick is to get GRT to create a set of parameters that do give you a correct value. When GRT is doing this iterative curve fitting process it is taking you rifle(chamber, barrel, bore, bullet, case, powder, primer) as a whole system. The option for OBT is where this iterative process is run. If you compare the original screen to the OBT screen you will see 2 parametrs are different. Those are:
1) Isentropic exponent (k)
2) Combustion coefficent (Ba)
Getting these right then gives good answers to where the nodes are. So here is the way I have found to give the shortest and most efficent process for a good load. These is where you can do it in 50 bullets.
1) Build your system. Get as much into the given sections as possible. Fired case water volume is important.
2) Run an OBT and look for a node that is at at about 85% of you case pressure.
3) Use that node you just creaated to load 8 bullets to the parameters that are in GRT. this will be the first iteration of determining those 2 parameters. Fire all 8 taking Velocity on them. I use the first 3 to foul the barrel and then the last 5 for velocity data.
4) Open your original base file and add to the "Measurement" tab the 5 velocity values you obtained. Now in the base file set up the powder steps so that the value you used for powder in thewse 8 bullets is created. Then select that powder level in the "Results" window. This then sets the system to do an iterative solution with one known point. One is better than 0 when we started but is not good enough to get an accurate projection of a node.
So run OBT now at this powder level. Figure out which node you want to aim for and then have OBT calculate that node. As a final step to document the optimized results hit the "create a OBT file". Once that file is created then cut and paste these to 2 values back into the base file (iteration 1 is now done).
5) For iteration 2 we will now build 20 bullets. Here I take the the node predicted in 4) as the centerpoint of 5 test powder values .2gn apart. I load 6 for the lowest value and 3 for each value above that. The first 3 are to foul the barrel. Then I take and shot these on a target that has 6 bulls so I can see the placement of each set relative to each other. Taking velocity for each shot fired.
6) now we go back to the base file and put these all in the measurement tab as before. I typically over write the first group values unless that value was within the new set of 5. Now I adjust the pwder spread to have a value that was the center value of the 5 sets and select that value Now run OBT again for probably the same node we used in 4). Create and OBT file. Back annotate the 2 parameters for iteration 2 into the base file.
7) At this point I am usually done. Usually one value on the 5 targets is close to iteration 2's predicted node, the SD will be low. Some times it is between 2 of the steps and if we are really blessed those 2 values have low SD's and you can split your new derived node.
8) Now we have come to the "proof of the pudding" I now load up 20 at this new value and go shot 4-5 shot groups. If the average is what I am looking for, I am done.
At this point if I haven't come up with a good load but the sd is still low, I then try seating depth to tune it in. So far I haven't had to try that.
So we used 8 + 15 + 20 = 43 bullets and 3 range trips to get a working load with an unknown system to GRT.
Once I had the powder values ( RL-16 and Varget) for my system I now plugged in those values to a different bullet and used that as step 4 results to do a spread of 15 bullets. And yes when I run OBT and get the values again they are slightly different but much closer than step 1.
Hope this makes sense. Unfortunately my main computer is a MAC and I have a Windows computer in the garage where I run GRT but getting screen shots into this MAC is cumbersome. But I think that if you are a novice to GRT once you have gotten through steps 1 and 2, the rest will flow easily.
David
Last edited:
