From the weak sauce that is the Gleipnir comes this weak sauce. In the small rope I'm using, this doesn't even hold. Looking at the final image leads me to believe that you might need some clarification of what is being done in the previous step.
Successes (holds ferociously in these, not backing off from the maximum tension I can apply):
1 mm cotton laid string;
soft 3 mm nylon braid;
firm and solid but flexible 5 mm nylon(?) braid;
hard stiff 5 mm kernmantel static accessory cord;
550 paracord;
soft to touch, firm to squeeze, but very flexible 10 mm kernmantel nylon braid with unknown linear core;
smooth, round, firm, flexible 1.6 mm nylon braid;
Failures: (backs off to some degree from the maximum tension I can apply): Stiff 10 mm kernmantel climbing rope with moderately high-friction surface, holds a kg-f or two of tension, but backs off to that level, moving by about two cm, after I release the maximum tension I can apply.
It will not hold at all, however, in any kind of cord, if the passes through the turNip are not in appropriately related directions. The direction for the first pass is arbitrary, but the second pass must enter the turNip by the same side by which the first pass entered it.
What material were you using?
I find the term turNip very useful for discussing this class of knots. I should not have lower-cased the N, thereby turning it into the name of a vegetable. I just edited the first post to correct this error.
To create the configuration in the second photo, I pulled to the right on the free end sticking up in the first photo.