your turning test should be done with a dumbbell shaped bar and with compound set to 5.75 deg. and then test cuts of tenths can be taken (3 or 4) then measurements taken.

If you are going by soft pull and hard pull thats not very scientific you should let the lathe do the work then measure and yes a lot of latje beds are flimsy to a small degree. the twist can be adjusted with adjustable feet.

What inserts are you using? The long stick out relative to the diameter (slenderness ratio) of 6.8 is well in excess of most recommended values, typically 2.5 to 3 depending on the material being machined. Coupled with the relative bluntness of the majority of carbide inserts it will result in the bar moving away from the cutting tool at the free end. This will give a larger diameter at the free end than is turned close to the chuck.

As Jason said, supporting the free end is a good starting point, then use ground inserts rather than just moulded ones, using a well sharpened HSS tool with a small radius tip to minimise the pressure on the bar is also better than most moulded inserts. Finally for long slim parts a travelling follower rest will reduce flexing of the part but is a lot of messing around if you can avoid it and does not work in all cases, shoulders can get in the way for example.

Martin C

Not at crossed purposes but debating if it is the fault of the lathe or simply bad practice with so much stickout and “blunt” tools that is causing the taper.

But how far is reasonable unsupported turning. I’d not want to start twisting my bed to compensate for what is most likely to be the work deflecting away from the tool.

Before touching anything I would try it with some 30-40mm bar and sharper cutting tools on a warmed up machine. Only then if the same amount of taper is present would I think about looking further.

Though we have not been given any indication of which way the taper runs which could eliminate the work deflection if the dia is smaller away from the chuck.