My flooring manufacturer requires no more than 1/8" over an 8 foot span. I have discovered that there is a fulcrum point along a crack in the subfloor that causes my 8 foot level to teeter. However, I am having a hard time determining the extent (i.e., width) of this bump by mere visual inspection. Should I just start by first (i) grinding the immediate area along the crack using my 4.5" angle grinder and then reapply the level to see if the condition has improved, and (ii) if the bump is still affecting my floor tolerance, expand the grinding to the left and right of the crack?

Moreover, is it necessary to place the level perpendicular to how the floor planks are to be laid? I had been checking the concrete by placing the level parallel to how the floor planks are to be laid.

Thanks in advance.

The only problem with grinding that way is your really guessing. It's very possible that the slab is heaving more heavily on one side of the crack, it can also change, along the crack, and be higher on the other side of the crack. It's pretty tough to predict rhyme or reason. Usually when one side of the crack shows an edge, that is the side to grind, but dont totally bank on that... The other side can be low I would suggest sliding the straight edge , perpendicular, up to the crack. This way you can see when the gap is starting ,at the bottom of the straight edge, as it approaches the crack. Mark that point, ( on the front of the straight edge ), along that high area ,for grinding reference.

The manuf. tolerance is not in reference to the direction you are laying the plank, but in ALL directions.

Floorologist,

I did what you said, and I noticed exactly what you had talked about (high/low spots to the left/right of the crack; no apparent rhyme or reason). However, it is obvious to me that along these cracks, I have issues (albeit slight in many cases).

What if I were to grind along the entire crack and then test the floor with the 8 foot straightedge? At the worst, I'll just have a depression along the cracks, which I can just trowel in feather finish along that narrow area. Is that a reasonable approach? It's either that or just giving up and applying slc over the entire area.

Your approach sounds good. Grind along the crack parallel to the crack, on the crack, and on both sides. This will flatten out the waviness along the crack. It's important that you grind parallel to the crack, otherwise your just adding to the waviness. Once you start grinding this way, you will visually see the the high and low spots as you scarify the concrete. Once all the high spots are down, it's a piece of cake to fill along the crack.

Howard,

I grinded parallel to the crack and about an inch to the left and right of the crack. However, there was still teeter tottering when I placed the 8 foot level at the center of the crack. Should I just use slc around this area?

You really just want to use filling compounds if you believe filling will ultimately bring the flatness within tolerance.
If you place the straight edge on top of the crack, perpendicular to the crack area, and the straight edge is teetering, or if you slide the front edge of the straight edge perpendicular up to the crack and there's a space under the straight edge, ( that's out of tolerance ), then there's still grinding of high spots needed.

Grinding a crack that's heaving in different areas can be tedious. Usually grinding 1" on both sides of the crack isn't wide enough. If you use the "front edge of the straight edge trick", you will be able to mark the points on both sides of the crack where that front edge starts raising. And grind the width to those points.

I think the "front edge of the straight edge trick" is eluding me. Do you mean that I should take the 8' level and start at a completely flat area that is perpendicular to the crack and then proceed to move towards the crack (while maintaining pressure on the level such that the level remains flat to the concrete) and then mark the spot where the advancing edge of the level starts to lift off the floor? Then I am to do the same thing from the opposite side of the crack?

Yes, that is what he means, and it is the quickest way to determine the extent of the high spot.

engineers catch on quick don't they?