I don't have any experience doing this myself, but I think your options are:

1. You could build or find someone with a bigger shear box; really big shear boxes or triaxial apparatus are occasionally referenced in the literature so I'm sure this is possible

2. You could use one of the many techniques for scaling the gradation curve down, all of which have limitations

3. You could try and build a large-scale field tilt test like reference in Barton Kjaernsli (Shear strength of rockfill), with the limitation that the normal stress will be very limited and so you are likely to end up with a quite high friction angle at low stress (angles up to 60-70 degrees are reported in the literature)

4. Yes, there is angle of repose, but angle of repose is dominated by 3D edge effects and possibly suction effects

5. Maybe some kind of in-situ direct shear test (not a tilt test) would be possible? Like, you dig out around a block of the placed material in the field and build a field appartus

The below links might help. I'd suggest looking for case histories / literature review for rockfill dams since I think that's one of the fields where this is been done and documented.

https://www.ecsmge-2019.com/uploads/2/1/7/9/217908...

https://www.issmge.org/uploads/publications/1/32/1...

https://www.researchgate.net/publication/298425776...

Angle of repose is the most practical way of doing it. We use type 6F material and 6P etc all the time. The whole industry adopts 45 deg so I wouldnt worry too much about it. You can do a sanity check by asking a lab to provide you with some large shear box test results (300x300). You will see that even 37.5m max dia aggregate is easily above 45 deg

Check the following paper: Soil improvement with vibrated stone columns — influence of pressure level and relative
density on friction angle (Wher 2008)

Although it is for stone columns, friction angles for rounded , smaller (2-32mm dia)river gravels have shone friction angles above 50 deg. For the type of material you are using, a larger, more angular aggregate, a design phi of 45 deg is conservative.

That wher paper looks consistent with what is seen in studies done for dams. Usually you have phi up to 70 degrees for low confining stress and 40-50 for high confining stress although most engineers seem to fall out of their chair when you tell them that.

recommend you look at the old US army corp reports and the 90's and 2000's work by Indraratna such as Large scale triaxial testing of greywacke rockfill.

They published results from large triaxial machines.

Self correcting myself here - maybe angle of repose would actually be the easiest way to do it. It seems to be angle of repose approx equal to PhiCV / PhiNC (or whatever variation of that concept you subscribe to).

angle of repose is the angel the soil sits in its loosest state. Get an excavator and drop a few good sized bucket loads from a height and let them naturally pile up. Measure the angle at a few locations. Take an average