Roll ultrasonic testing decides whether a forged roll leaves the shop. When the test fails, the cause usually sits far upstream of the test itself. High-alloy 70Cr3NiMo steel now covers hot rolling work rolls, cold rolling work rolls and backup rolls. Rejection traces back to two stages: steelmaking and forging.
High oxygen in the steel brings subsurface blowholes and porosity, and it sharpens the harm that sulphur does. Mature melting routes and newer forging methods exist. Aún así, the roll forging process stays complex, the factors that govern internal quality are many, and control is imperfect. Rejections still happen, and each one costs money.
The Case: Uno 10,500 kg Ingot
A plant cast a 70Cr3NiMo ingot weighing 10,500 kg on an EAF + LF + VD route. After forging, roll ultrasonic testing found dense defects and single-point defects above the limit. Their equivalent size failed the class III requirement of GB/T 6402-2008.
The test ran on an EPOCH 600 unit with a PF2R-24 probe, on the blank. Defects sat about 860 mm from the nozzle side. They formed a dense cluster at Φ3 to Φ4 mm equivalent, with a single maximum of Φ6 mm. The defect echo cut the back-wall echo by up to 12 dB. Depth stayed inside Φ140 mm at the roll centre.
How the Roll Was Forged
The shop did not cut the nozzle and riser off the ingot before forging. It forged the roll first and cut them afterwards, which put ingot utilisation at 72%.
Forming took two heats. Start-forging temperature was 1180 °C and finish-forging temperature 950 °C. In the first heat the press drew a tong hold, chamfered the ingot and upset it to roughly 800 mm high by Φ1280 mm. The block then went back into the furnace. After the second heat the press chamfered it, drew it to a 670 mm square and forged the roll to shape. One ingot formed two rolls, and the second roll followed once the first was cut off.
What the Section Showed
Dye penetrant on the longitudinal section
The shop cut the rejected roll along its axis and ran PT on the section. Five clear crack-like defects appeared, all near the roll centre. Two of them, A and B, stood out.
Stereo microscope
Samples 1# y 2# came from defect B and went under a Stem2000-C stereo microscope. Both defects ran along the main deformation direction and sat in bands, and both cracks were thin and straight. Sample 1# measured about 1.13 mm long, and sample 2# acerca de 3.37 mm.
Metallography
Metallographic samples went under an Axio Imager A2m. Part of the 1# defect had a smooth wall and a residual length of 610.69 µm, and non-metallic inclusions sat inside it. El 2# defect measured 8386.83 μm and carried a large amount of non-metallic inclusion as well.
EDS on the inclusions
A ZEISS SUPRA55 scanning electron microscope ran energy-dispersive analysis on the inclusions inside both defects. Oxygen and aluminium formed the main elements. Aluminium deoxidation leaves solid Al2O3, so the inclusions were alumina.
Why Roll Ultrasonic Testing Reads Inclusions as Cracks
The inclusions cut the continuity of the matrix. Forging flattens the inclusion zone and thins it, yet the press cannot weld it shut. What remains is an elongated “crack”. That feature is what makes roll ultrasonic testing fail. The indications are not cracks in the usual sense, and no heat treatment closes them.
Inclusions do further damage during forging. Where an inclusion will not deform, or deforms with difficulty, a gap opens between inclusion and matrix and becomes a void.
Where the Inclusions Came From
The steelmaking route
Three paths put alumina into the steel. Primero, primary deoxidation products stayed in the steel because they never floated out. Segundo, aluminium dissolved in the steel oxidised again while the steel poured in air. Tercero, residual aluminium kept deoxidising as the steel cooled and solidified in the mould.
All three leave the same marker: oxygen and aluminium together, in a string or a band, along the working direction.
The forging route
The second source is forging practice that does not suit the ingot. Because the nozzle and riser stayed on during forging, original defects at the nozzle end travelled into the roll body with the rest of the ingot. Utilisation at 72% left that metallurgical defect zone inside the part instead of in the discard. Drawing to round then spread it through the roll centre, which is exactly where roll ultrasonic testing reports the indications.
Five Process Changes
- Stop slag in two steps during tapping. Less carry-over slag means less Al2O3 in the ladle.
- Control the argon stir. Keep exposed steel at or below 300 mm, hold bottom-blow flow between 20 y 100 N·m3/h, and after feeding aluminium wire allow at least 5 minutes of soft blowing so the alumina can float.
- Treat in the LF according to the argon station sample. Deoxidise and desulphurise, then feed CaSi wire for calcium treatment. Calcium modifies manganese sulphide and alumina, which cleans the steel. Cutting sulphur and oxygen also suppresses type A and type D inclusions.
- Protect the cast. Shielding the stream against air removes the second oxidation path.
- Cut the nozzle and riser before forging. This keeps riser and nozzle defects out of the body. It also means accepting a sensible utilisation figure, because a high yield is exactly what leaves those defects in the part.
What Changed After the Fixes
Sampling pointed to two causes at once: non-metallic inclusions above the permitted class inside the ingot, and forging practice that kept them. With slag carry-over controlled, alumina reduced, and the forging sequence corrected, the forged rolls met the acceptance requirement for roll ultrasonic testing.
ESR steel goes a step further. Electroslag remelting raises raw-material cleanliness and improves roll performance, and it cuts the share of rejections that come from inclusions over the limit. For cold rolls in particular, that route is the practical answer when plain ingot practice keeps missing the UT class.
Conclusión
For this rejection, sectioning showed super-scale non-metallic inclusions in the ingot plus a forging route that retained them. After slag control, lower alumina in the steel and a corrected forging sequence, the forged parts met the roll ultrasonic testing requirement in full.
