I have a set of 906 casting Vortec heads, are these heads cracked? want to find out? I do before I put them on my engine and and try and drive well, anywhere... Now, the how to, and with what.

There's a few ways to test for cracks, in this tutorial, we'll be talking specifically about dye penetrant(DP) testing. DP is performed by spraying the part to be checked with a dye, and letting the dye soak into the part, then removing all of the visible dye from the part, and covering the part with a developer, which will show the slightest bit of dye you did not remove. the dye will work it's way into the tiniest crack you can imagine, so as you may imagine, it can show you more than just cracks. things such as tool marks, surface porosity. joints, weld edges, ect, will all hold dye, so it's important to make sure when you have an indication, that you understand it doesn't mean there's a crack, but the indication needs to be carefully inspected to make sure that it isn't.

I am using Dynaflux Crack check, it's about $150 for the kit, so it may or may not be cheaper to have you machine shop do a magnaflux depending on how much testing you need to do.



some notes before you start, you want to do this inspection in an area that you're not worried about getting messy, and you most certainly want to wear gloves, and protect any clothing you don't want ruined.

it should make sense that we want to start with the cleanest, most defect free surface possible, to avoid possible false indications. All dirt, oil, grease, and paint needs to be removed from the test area, and the test area should be dry. shake the dye, and evenly coat the entire surface of the test area, and let the dye sit. the Dynaflux instructions say 1-30 minutes, the longer you wait, the more indications you are likely to get. it's important to note, the dye needs to cover the entire surface, almost pooling on the surface.



Next, you need to use a clean, ideally lint free cloth to wipe as much of the visible dye off as possible, this will probably take a while, but you want to go at it pretty hard to minimize false indications, focus on things like weld toes, joints, and sharp angle changes, these areas will all hold dye very well.

at this point, sometimes you will already have crack indications, while I wouldn't call it without taking a few more steps, look carefully and see if you can spot any dye oozing anywhere.



This is a pretty sharp indication, also notice how it's on both sides of the hole, lending a bit more credit to it probably being an actual crack.

now that the dye has been removed from the surface, shake and spray on the developer, with the dynaflux kit, the developer is a white absorbent material that will wick the dye out of the crack, or wherever it's hiding. note that just because the developer has changed color, doesn't mean there's a crack, look at the valve seats, or the non machined parts of the casting, the dye pooled there and is not presenting itself, but it's not a crack.



now, the crumby part for me, is that spot right in the middle, yep, still looks like a crack... at this point, I have several indications that this is a real, legitimate crack.
-Dye present after wiping surface,
-Dye present on both sides of the bolt hole
-Dye showing after wiping a second time and spraying developer.

but for demonstration purposes, lets say this was in the middle of the surface, and maybe there was a small machining defect there, how do we challenge the result? there's 3 ways, I check the validity of the results

1. wait. a true crack will continue to weep into the developer, and the lines made by the dye will get wider.
2. say it's not getting any bigger, give it a short burst of brake clean, look in the area, if it's a crack, dye will likely be visible weeping.
3. retest. in this case, I usually start by cleaning the area indicating, and spray more developer on, if there's a crack, it will show.



Cleaning for retest note the dye weeping on the near side of the hole


more developer, note the crack is visible in two directions, both across the face, and down into the bolt hole


Here, you can see the dye causing a larger indication after some time passed.

Now, we can see there's definitely a crack, the next step here is to evaluate the significance of the crack, is the crack in a location where it doesn't matter? Is the crack a common non-issue? what could happen if the crack gets worse? In this case, the crack is in the water jacket, to the cylinder head bolt hole, it's possible for this crack to leak coolant into the bolt hole, and then to the area under the valve cover, and then to the oil. it's also possible it doesn't leak at all. I could sleeve the bolt hole, but I'll probably replace the head, because this wasn't the only crack I found unfortunately. This process works for finding cracks on more than just heads too, got a leaking pipe? want to test a weld? have other possible indications of a crack like flaking paint? it's a fairly easy process, and not too time consuming.

Hopefully your crack checks go better than mine!