You have not shown any code, if you want people to help you solve this problem you should post the part with the code that is not working for you.
My guess is that your vertex normals are not pointing outwards properly at the angle they need to be and there for raytrace system cannot see them maybe because of this. You should also look at the ray trace system because it has an ignore list and an allow list, and maybe the normals are on these lists or are not.
So the cube vertex normals if it’s just a face of a cube with just 4 vertices per face then the normals will not maybe point in the direction of the raytrace system because they are ajusted at 0.5 as I told you, they will not point in front of the cube maybe, forward, they will point halfway upwards.
These are soft normals and will not point towards the raytrace.
I still don’t fully understand, you tried to get normals and you could not get them or you don’t know how to get them.
You should try PMC, it has all the data open once it becomes a procedural mesh, and to your trace system you can specify variables that are open inside the procedural mesh system.
For procedural system the normals has a whole array you can call and you can specify with your trace system depending how you have it set up, If trace system hits “var” whatever then do whatever.
So everything opens up in PMC and becomes availible
This is where the normals are stored for example.
TArray < FVector > normals;
So as you can see the vertex normals are a vector.
If you have a mesh sections, perfect, you load the whole mesh section into these kind of arrays with this library. UKismet library it will load all the mesh things, tangents, normals, etc and now you got easy access to normals with your trace system.
It’s possible.
Review this thread.
So from what I can see the user here ray traced the normals from what I see.
Examples:
In this simple example I generate 4 blocks on top of each other from bottom to top (randomly picked “rock” or “dirt”). The floor is a chrome plane for reflection. As you can see, the second Mesh Section (“dirt”) has correct Normals in “World Normal” view, but grey(?) Normals in “Ray Tracing World Normal” View. Thus, the reflections in the left image are somehow missing. SSR work fine though (probably since the normal pass is correct).
It’s also possible with blue prints to raytrace stuff in PMC.