Model Engineers’ Workshop – August 2019

68 http://www.model-engineer.co.uk Model Engineers’ Workshop

1. The pressure exerted by the rolls to
   convert the square section to half-round
   is transmitted throughout the section of
   metal being squeezed. The incompressible
   metal behaves rather like a very viscous
   liquid, and instead of remaining
   4mm wide, it oozes out of the
   unsupported sides resulting in the
   cross section shown in photo 10.
   If I’d been working with copper,
   or some other inexpensive metal,
   this would have been easily
   remedied by simply fi ling away the
   unwanted extrusions. But precious
   metal is far too expensive to waste
   in this way.


2. As a result of there being
   no lateral support in the gap
   between the rolls, the pressure
   on the twisted part causes it to
   partially untwist, resulting in rather
   distorted hoops in the fi nal ring.
   This is totally unacceptable.
   The only solution I could think
   of was to make a new pair of rolls
   with supporting sides, as shown in
   photo 11.
   Since I had two rings to make;

one 4mm wide and the other 3.5mm wide,

I needed two pairs of special rolls. The

prospect of machining 7mm deep, half-

round grooves into steel bars was not that

appealing, so I decided to make them in the

form of separate narrow inner rolls that use

the same pair of steel cheeks. Photograph

12 shows the two pairs of inner rolls and

the steel cheeks. One pair of inner rolls is

sandwiched between the outer cheeks by

means of high-tensile screws that screw

into threaded holes in the left -hand

cheeks.

Since the main rolls are 65mm in

diameter, both pairs of inner rolls had

to be 65mm too. That meant that one

pair of cheeks had to be 75mm and the

other only 55mm. The next job was to

make them.

The diameter of the roller axles is

28mm, and the width of the original

roller extensions is 20mm, which means

that the width of each cheek must be

8mm. I had two bars of mild steel (MS),

one 75mm and one 65mm in diameter. I

fi rst bored a 22mm deep, 28mm hole in

the 75mm bar, and another one, 50mm

deep, in the 65mm bar. I then faced both

bars at both ends and mounted the

75mm one vertically on the mill table,

and drilled three equally spaced, 22mm

deep 5.3mm holes, on a 39mm PCD

around the centre.

The bar was then returned to the

8

9 10

11