3573, 7249, and 3572

[Peter Blomberg]

The arch 3572 may possible match the curvatures of the sloped bricks 3573, 7249, 50950, and 24309.
We need to confirm these matches using physical parts.

Does anyone have two or more of these parts?

A. Does the curvature of 3573/7249 match the curvature of 50950/24309?
B. Does the curvature of 3573/7249 match the curvature of 3572?
C. Does the curvature of 50950/24309 match the curvature of 3572?

I know what BI data claims, but now we need physical confirmation.

[Orion Pobursky]

For reference:
3572: https://rebrickable.com/parts/3572/brick...1-x-5-x-2/

3573: https://rebrickable.com/parts/3573/slope...-no-studs/
7249: https://rebrickable.com/parts/7249/brick...4-x-1-2-3/
50950: https://rebrickable.com/parts/50950/slop...-no-studs/
24309: https://rebrickable.com/parts/24309/slop...-no-studs/

[Cam's Bricks]

You can find your answers here
https://www.newelementary.com/2023/07/le...rolet.html

and here
https://www.newelementary.com/2025/05/pa...fetts.html

Also the 3 long slopes (50950/24309) are confirmed for years so I cant find a refence image but they definitely are the same. I really recommend following the New Elementary blog as a great source for how the system works.

[Peter Blomberg]

Awesome!

Now that we know they all have the same curvature, we need to find out which is modeled inaccurately. Is it the 50950 or the 7249?

Does 50950 have a matching arch?

[Peter Blomberg]

So, I did some measurements and it seems 7249 is correct and 50950 is 1 LDU too low in the front.
   
Can anyone confirm these findings?

[Rene Rechthaler]

the problem is that 50950 has a lot of files (stickers) and its family (24309, 78522) too...
Should these be the same curvature as the big curved slopes?
(50950 has no top prim, I fixed that but its not uploaded yet...)

[Orion Pobursky]

the problem is that 50950 has a lot of files (stickers) and its family (24309, 78522) too...
Should these be the same curvature as the big curved slopes?
(50950 has no top prim, I fixed that but its not uploaded yet...)

50950 has 21, 9 of which are unofficial. The error is so small that I think we can fix these without obsoleting.

[Philippe Hurbain]

the problem is that 50950 has a lot of files (stickers) and its family (24309, 78522) too...
Should these be the same curvature as the big curved slopes?
(50950 has no top prim, I fixed that but its not uploaded yet...)

Maybe we could correct the top curve of 50950 by slanting a bit the top surface? Then all the stickers could be easily updated by applying the same slant. Still a LOT of work, but maybe feasible?

[Gerald Lasser]

Sorry for chiming in late here, I still have a HOLD on the part 3573 as the curve is not correct, and needs to be slightly higher

NewElementary is actually not correct here, the curve is not a continuation of the 50950, see the picture of my real parts here:
   

Here is a digital comparison:
- Yellow/Red: Ldraw
- Blue/white: BI-raw data

   

In my view:
- the curve of the 50950-family is pretty accurate to the BI Data
- the curve of the 3572/3 is too far of the BI-model and needs to be adapted to be closer to the BI curve  
     

Here is a comparison including the LDraw 7249 which is closer to the BI-Mesh and which represents the picture I took
   

Both share the same shortcoming that all of the curve slope and slope parts have, the 4 LDU hight lip. The lip is 5 LDU in the BI meshes, which I would not touch

So my way forward would be:
- Keep the lips at 4 LDU
- Keep the 50950-family
- Adapt the curve of the 3572 and 3573 to the 7249

BR

[Cam's Bricks]

very interesting. thanks for correcting us!

[Magnus Forsberg]

Uploaded corrected shapes now.

[Rene Rechthaler]

can we still synchronize the 50950 5-face curve to the 24309 6-face prim?
(would still mean to redo the stickers)

[Peter Blomberg]

So my way forward would be:
- Keep the lips at 4 LDU
- Keep the 50950-family
- Adapt the curve of the 3572 and 3573 to the 7249

Thanks Gerald!
Much appreciated. I fully support this way forward.