This article is a reverse engineering of a certain mold change of part Slope 2 x 1 with 2/3 Cutout.


The idea for this article came to me when recently I was browsing through my inventory on Rebrickable and noticed that there are two part numbers (aka design IDs) assigned with the aforementioned slope:

  1. 92946 - This is the original version of the part that was introduced in 2011 and was designed as a modification of the long-lived Slope 2x1 (3040b).
  2. 15672 - This is a mold variation appeared in 2014.

As you can see, the part itself is fairly new so it may be quite interesting to learn what is the difference between the two versions and what was the reason for the mold change. I could not find any information on the subject so I tried to figure it out myself. Hence the following observations may not be absolutely accurate and all conclusions should be treated as my personal opinion.

In this article I will be using parts with the following colors (refer to Pic.1):

3040b - Dark Bluish Gray; 92946 - Blue; 15672 - Light Bluish Gray.

Part numbers are imprinted on their underside as shown on Pic.3.

Visual examination

At first glance, both versions look identical. Undersides are slightly different (refer to Pic.2) but the difference is negligible. Nevertheless, part numbers imprinted on the inclined walls of the parts (Pic.3) clearly state that they are different molds.




However, if we put the slopes side by side, it becomes evident that while 92946 is exactly 1 brick tall, 15672 is slightly shorter (refer to Pic.4). Moreover, the inclination angles are also different (Pic.5): while 15672 is true to its original 3040b (Pic.6), 92946 is slightly steeper (Pic.7). To summarize, both versions look somewhat imperfect: 92946 has different inclination angle than its predecessor 3040b, and 15672 is not exactly 1 brick tall.

What is interesting, current version of LEGO Digital Designer (version 4.3.8, brick set 1564.2) acknowledges the difference between 3040b and 92946 (refer to Pic.8).






If the above differences still look insignificant to you, check out the picture below. As you can see, 92946 may fail in assemblies which are perfectly normal for its predecessor 3040b. Once again LDD acknowledges this fact: the first assembly from the photo is allowed in the program while the second one is not. All in all, part 92946 has a significant design flaw which may be the very reason for the mold change to part 15672 which is free from this issue.


By the way, since I have already mentioned the inclination angle, it would be interesting to check if these slopes live up to their family name "Slope 45° 2 x 1". Here part 92946 is at its best with an inclination angle of exactly 45° (Pic.10) while the inclination angle of 15672 and 3040b is notably smaller (Pic.11).



Two questions still remain: why exactly the middle assembly on Pic.9 does not work and why 15672 is less than 1 brick tall.


Before I can move to measurements, I should mention one important property of LEGO bricks — horizontal tolerance.

Horizontal tolerance is a tiny gap subtracted from every vertical wall of the bricks to prevent friction during building process. It is believed that the horizontal tolerance is always 0.1mm. In other words, while the distance between each two adjacent studs or antistuds is always 8.0mm, every real brick or plate is shorter and narrower by 0.2mm than its 'theoretical' size. E.g., brick 1x2 has a footprint of 7.8x15.8mm (instead of 8x16mm). Pic.12 demonstrates the gap between two bricks connected to other bricks. You can read more information about the horizontal tolerance and brick dimensions in a series of articles by Robert Cailliau.

Horizontal tolerance may seem absolutely harmless, but apparently its existence has lead to some controversial decisions when designing SNOT elements (bricks with studs on sides, brackets etc.). This results in a possibility of SNOT assemblies that seem perfectly legal in theory and in virtual editors, but fail in reality. Two examples of such assemblies can be found here and here.


Now I can proceed to measurements although I should note that I usually try to avoid using 'human tools' for measuring LEGO parts and instead compare unknown parts to parts with certain dimensions. Some of my measurements may not be absolutely precise.

First of all, we need to measure the original part 3040b. It is quite obvious that its top and bottom sides are equal in size to bricks 1x1 and 1x2 respectively (refer to Pic.13). However, as I have already mentioned, the width of brick 1x1 is 7.8mm which is smaller than half of the length of brick 1x2 (15.8mm). Hence the horizontal size of the sloped section is 8.0mm.

Now we need to determine the height of the bottom lip of the slope. Despite the assumption often used by builders, Pic.14 clearly demonstrates that the lip is taller than half of a plate i.e. 1.6mm = 4 LDU (LDraw units). This is quite understandable considering that the underside of the slope must accommodate a stud which has a height of 1.8mm (once again refer to the article by Robert Cailliau).

Assembly shown on Pic.15 helps to get a better approximation. Thin Light Bluish Gray parts under the slope are neck brackets which are believed to be 3 LDU thick. Together with the lip of the slope they are equal to the height of a plate i.e. 8 LDU. Hence the height of the lip is approximately 5 LDU = 2.0mm.

The dimensions of part 3040b are summarized on Pic.16. The sloped section has a length of 8.0mm and a height of 7.6mm which gives an inclination angle of approximately 43.5°.






Now I move to the slopes with cutout. Pic.17 demonstrates that there is a tolerance gap between the vertical walls of the slopes and plates connected to the cutout section. It serves the same purpose as the gap shown on Pic.12 - to prevent friction between bricks. It is also evident that the sloped sections of parts have footprints equal to brick 1x1.


The probable reason for the whole mystery is that the need for the aforementioned tolerance gap does not allow creation of a slope with cutout which would have the same height as its predecessor 3040b. 92946 and 15672 represent two different approaches to this problem: while the slope of 15672 goes at the same angle as 3040b and does not reach the top (Pic.18), the slope of 92946 goes steeper to compensate the gap (Pic.19).



Pic.20 shows a schematic comparison of all three slopes (note that the tolerance gap in the middle is exaggerated while the width of the top edge of slopes 92946 / 15672 is ignored).

As for the failed assembly from Pic.9, it is now quite easy to explain - refer to Pic.21.



To conclude, two versions of Slope 2 x 1 with 2/3 Cutout may be characterized as follows:

  • Part 15672 is truly a modification of 3040b as it has the same inclination angle and can be used in the same assemblies as its predecessor. Its only disadvantage is smaller height which leads to a gap above the top edge of the slope. It may have a negative aesthetic effect because the parts behind the slope are more exposed to view.
  • Part 92946 is rather a slightly different design than a direct modification of 3040b. It has a correct height of 1 brick and a correct inclination angle of 45° but it may fail in assemblies that are perfectly normal for 3040b and 15672.

Actually it seems quite strange that part 92946 was released in the first place because it breaks the rule of backward compatibility of parts, and it is reasonable that it was eventually replaced with 15672.

The article comes to an end. If you have any corrections or additional information, feel free to post a comment below. Thank you for your attention!


  • 1 year, 3 months ago dna2 (94)
    Great artical, thanks! Yes, more please!
  • 1 year, 3 months ago MocJens (670)
    really good article.. first time I read an article here on rebrickable an love it :)
  • 1 year, 3 months ago Kalais (1239)
    Please do more articles like this :) I love it :)
  • 1 year, 3 months ago boneskull (1969) MOC Designer
    Guys, this was amazing. More like this.
  • 1 year, 3 months ago Immo (3494) MOC Designer
    Thanks for that, very thorough analysis :)
  • 1 year, 3 months ago zux (1836) MOC Designer
    Post about parts and their specific differences? Yes please, I could read this any day. Hope it is not the last one.
  • 1 year, 3 months ago chrisfranklyn (20)
    I've been scratching my head over this just over the past days, so the timing of this is perfect. Thanks!
  • 1 year, 3 months ago Late4Dinner (120)
    Excellent observations, and a very well written post thanks ☺
  • 1 year, 3 months ago TobyMac (31539) Inventory Admin ADMIN
    Thanks for the interesting post. Please let them keep coming (as that a correct English sentence?)
    • 1 year, 3 months ago chrisfranklyn (20)
      It's not wrong English, I suppose. :) But people would generally say "Please keep them coming"
  • 1 year, 3 months ago biodreamer (18477) MOC Designer
    two down 100000 more to go ;) keep up the good work
  • 1 year, 3 months ago Bolbuyk (8529) MOC Designer
    Nice study of this part.
    I like that!
to your account to add or reply to comments.