This flaking is happening at the insight-live.com/glossary/261">interface between the body and orange commercial underglaze (the firing is cone 05). The reason is simple.
Pure stains contribute only one ceramic property: Color! Commercial underglazes sold in jars dilute stains into a recipe of materials, a powder medium, designed to impart the missing properties. However, suppliers like to quote products as suitable over a wide range (e.g. cone 06-6). Of course, at cone 06, they have almost no glass development for bonding to the body and at cone 6 they may well be melted. In this case, the stain used is refractory and a high percentage is needed. So the bond with the body is very poor. A thick layer coupled with an incompatible thermal expansion can make the underglaze/glaze layer pop right off.
Is a DIY version possible? Imagine if you created a powder medium (e.g. L3685Z3) containing enough frit to counter the refractory character of the stain and develop the needed glass to bond with the body. That would solve this problem.
Pages that reference this post in the Digitalfire Reference Library:
Creating a Non-Glaze Ceramic Slip or Engobe, Pure stain powders painted overglaze: Why the bleeding and bubbling?, An engobe is shivering off the rim of a low temperature mug, Make AMACO velvet underglazes work better at cone 04, Thickly applied slips must fit the body and each other, A reason for DIY underglazes: Commercial products don't adapt the medium, Stain Medium, Underglaze, Body glaze Interface, DIY Underglaze Development

This post is one of thousands found in the Digitalfire Reference Database. Most are part of a timeline maintained by Tony Hansen. You can search that timeline on the home page of digitalfire.com.