Lathe Rigidity Issues – Modification Opinions

[Ed PageParticipant @edpage45070]

Hi,

I'm no stranger to the forum, although not logged in for such a long time it seems I needed a new account.

I live in Canada which seems to have no market for used small lathes like the UK did, and I didn't have the room for a large lathe, so I had to purchase one new and convert it to CNC. I am a machinist by trade so the lathe is a little under-whelming. I also have a 4 axis CNC mill to make some bits on if I have to. I no longer have access to any other machining tools apart from this lathe and mill.

This lathe is the Grizzly 0602 equivalent, a pretty generic lathe design with lots of different names to it. I first removed the compound and replaced it with a new holder, made from 4140 to hold 1 inch standard lathe tools, also to make it rigid as possible. I converted it to CNC, it is running some high torque servos, runs at up to 10m/m, so it's fast. However it is not very capable in terms of rigidity. I have been making some pistons out of 1018 steel which I managed to turn at 0.4mm/rev and 10 thou DOC, that was the limit of the motor, but also shallower cuts caused vibrations. I could only hit 3 thou before the CNC conversion, it now maintains half a thou. It finally got to cutting the grooves in the pistons to find all it would do is chatter, in the end I used my 4th axis on the mill to do the grooves, but that was quite time consuming. The piston is 5 inch in diameter.

I have tightened the gib on the X-axis as tight as it should be, there is no adjustment on the saddle, the weight of the servo actually lifts it off the back ever so slightly. There is a small strip of steel bolted at the back underneath the saddle to keep it from lifting, without that the saddle would literally fall of the front. I know there are lots of solutions but I would like to know your guys opinions. Here are a few of my ideas in order of complexity.

1. Bolt a 20lb slab of lead on top of the X slide, that should dampen resonance and the weight should take out any slack. However not sure that will take slack out of the X axis, and it may put a lot more wear on the bed.

2. Mill the saddle flat, hopefully its not hardened. Bolt some linear rails on, make a new X- slide out of a chunky piece of steel. I never need to turn any more that 6 inches, and I do have a 2 x 6 x 12 piece of flat bar left over.

3. Make a completely new saddle and x-slide. Line the saddle with either plastic or bronze against the bed surface, make it very heavy, 100lbs or more total.

[Ed PageParticipant @edpage45070]

[Ady1Participant @ady1]

The piston is 5 inch in diameter

lol

Its never going to be easy to do a huge lump like that on a hobby lathe

No real stiffness so high speed and sharp carbide

The real solution if you're doing a lot of work that size is a bigger lathe

[Dave SParticipant @daves59043]

My first thought looking at the photos is that your grooving tool has a lot of stickout.

If your saddle isn’t in contact with the bed properly I suspect there is a

potential for “bounce” for want of a better description – the saddle dips as the cut comes on, then as the chip peels it lifts, changing the tool angle, which causes it to dip again….

Can you add ball bearing gibs like on a 10ee to the back of the saddle?
Dave

[JasonBModerator @jasonb]

You could also try a narrower tool. You say you could only manage 10thou DOC when turning so as you are now using a grooving tool that is maybe 100thou width of cut that is putting a lot more load on the light hobby machine. Something like a 1mm wide carbide insert of the "Minithin" style would be how I would do it on a manual machine at those diameters.

Servo at the back may balance out the apron at the front.

Some people also change the plate that held the topslide to a rectangular one with a fixing in each corner and say it helps with rigidity, looks like you have retained the original plate.

Edited By JasonB on 18/03/2021 08:29:09

[Nick WheelerParticipant @nickwheeler]

Posted by JasonB on 18/03/2021 08:26:43:

Some people also change the plate that held the topslide to a rectangular one with a fixing in each corner and say it helps with rigidity, looks like you have retained the original plate.

I could see the original plate on WM250 bending when parting off or making heavy cuts. I replaced it with a 15mm thick plate, using all of the original topslide fittings and four M6 t-nuts. The original plate needed to be done up really tightly for any clamping effect. The new one just needs the four bolts to be nipped up with a 10mm spanner for much better clamping.

The new plate was hacked out a larger chunk with an angle grinder, the cut edges filed smooth, and the mounting/clamping faces machined in the lathe. I keep promising myself I'll paint it, but that's unlikely to happen until I paint something yellow….

I can't imagine using 1" tools on such a small machine, and definitely wouldn't want to pay for them. Most of mine are 8mm shanks, used in a Myford size Dickson QCTP, and they're not the limiting factor.

[Speedy Builder5Participant @speedybuilder5]

A bit of a job now, but put the stepper motor on the rear of the saddle ??

Oops – just saw Jasons comment.

Edited By Speedy Builder5 on 18/03/2021 09:03:05

[Anonymous]

Insert blades have a propensity to chatter unless the feed per rev is above a minimum value. I use inserts for parting off with a minimum feedrate of 4 thou per rev. But for grooving I grind a HSS tool bit which will give a better finish and no chatter even when sneaking up on a dimension.

Andrew

[HopperParticipant @hopper]

If I was doing that job in my Myford — which looks to be a similar sized lathe — I would be using back gear to spin that 5" piston at less than 100rpm somewhere. Will your Chinese lathe run that slow without losing torque?

And I'd reduce the stick-out on that grooving tool to the absolute minimum, like the groove depth plus 1mm. The round toolpost has the disadvantage of curving back the wrong way right where you need max support on the cutting edge side of the tool unfortunately. A grooving tool with the cutting edge on the right-hand side of the tool would be better supported.

Fort the wider grooves in the final pic, I might use a regular left-hand and right-hand turning tools to make that. Cuts nicer than a grooving tool when moving it from left to right to widen the groove. Just have to use the two different turning tools to get the square shoulder at each end.

And I find that in rigidity-challenged situations such as grooving on these small lathes, a razor sharp HSS tool can chatter less than carbide.

[Tony Pratt 1Participant @tonypratt1]

Worth trying a shorter grooving tool, easy & cheap. I found on my Myford super 7 the revs had to be way down low to stop chatter like this.

Tony

[Niels AbildgaardParticipant @nielsabildgaard33719]

The lathe is a 26mm bore 250 size like the much loved one I had before my son liberated it.He will get a surprice when estate is parted.

First thing is that it shal have a solid foundation.

The old Emco toolholder system is idiotic and a momemade came on

[Niels AbildgaardParticipant @nielsabildgaard33719]

For parting on my Boxford I made a very stiff system and for my 250 with solid flange another.

The ME photo insert system is a disater waiting to happen for us unpaid writers

 

Edited By Niels Abildgaard on 18/03/2021 10:52:13

[Dave HalfordParticipant @davehalford22513]

Firstly skim the plate seating (not the saddle) on the back of the saddle so you have minimal clearance, that will stop the saddle lifting and dropping and lifting ——- in time with you chatter. This is due to the front being forced down by the cut which can't move and so the back is lifted instead, which raises your tool height, and stops cutting.

If you throw excess weight at the saddle the lathe bed will begin to bend.

I would make the tool tower square. If you look at the tool / tower contact patch that you have it's also quite small.

 

Edited By Dave Halford on 18/03/2021 11:34:16

[Tony Pratt 1Participant @tonypratt1]

If people are looking for 'stiffness' in their tooling systems why do some many have a tendency to use round tool towers etc when square would be so much better?

Tony

[JasonBModerator @jasonb]

This guy has a nice solid replacement for the top slide

[Ex contributorParticipant @mgnbuk]

The undersides of the bed front and rear are bearing sufaces for the keep strips that constrain the saddle vertically. These faces should be machined parallel to the top bed formation to a similar standard of finish. The keep strips can be solid blocks, ground or scraped to be in running contact with the bed when bolted up tight, or a bolted on piece designed to hold a tapered gib strip for easier adjustability. The saddle should not be able to lift regardless of how the load is applied – adding weight to the saddle to keep it in contact with the bed is not the right way to address the lift issue. IMO correctly adjusted saddle keep strips are an essential part of the machine.

Changing to linear bearings would probably make the situation worse rather than better – they make for very "lively" axes where what you want is better damping.

The bed design on these doesn't seem too good if it is similar to the model Nils shows – in one of his other posts Nils showed one of these machines stripped down & I was struck by how poorly the headstock is supported due to the amount of metal removed underneath it to make space for the motor. Most Western lathes continue the bed at it's full form (or beef it up) under the headstock to properly support it.

If I was to add weight it would be in the form of a substantial flywheel mounted at the rear of the spindle.

Tool overhang, cut width, toolpost footprint, direct drive variable speed spindle drive shortcomings & the requirement for a solid foundation have already been mentioned above. I would also look at adding non-drive end support bearings to the ballscrews & really the Z axis arrangement is "the wrong way round" – the motor & end support bearing bracket would normally be positioned where the feed gearbox is, so that the ballscrew is in tension when under normal cutting loads & so not subject to buckling, with a support bearing at the tailstock end. On a screw of that length, I would have a bearing arrangement that allowed the screw to be tensioned rather than just a single deep groove ball race support. Hiwin Ballscrews application notes give details of how ballscrew performance is affected by support bearing arrangements – a copy can be found as a .pdf on line.

If it is any consolation, even large industrial lathes can struggle with grooving. At my last employment we sold a straight bed Boehringer CNC lathe to an aerospace company & I was sent over to investigate when they complained about chatter when grooving. The machine was (IIRC) 800mm swing x 4 meters between centres, 22Kw drive with 4 gear ranges & probably weighed in at around 16 tonnes, but it chattered when the customer attempted to use an 8mm wide grooving tool. Tried all sorts of mechanical adjustments, control and drive settings but to no avail. Cutting the groove with a 4mm wide tool was chatter free, but the customer was adamant that the part had to be made with the 8mm item & the machine, in the end, just could not do it.

Nigel B.

[Michael GilliganParticipant @michaelgilligan61133]

Posted by JasonB on 18/03/2021 13:09:26:

This guy has a nice solid replacement for the top slide

[…]

.

__ I would probably even be tempted to tap the LIKE Button if we had one !

MichaelG.

[JasonBModerator @jasonb]

Posted by Michael Gilligan on 18/03/2021 14:36:21:

Posted by JasonB on 18/03/2021 13:09:26:

This guy has a nice solid replacement for the top slide

I would probably even be tempted to tap the LIKE Button if we had one !

He also has a very nice home made epoxy granite CNC mill, have a look at his "Elke" videos. Youtube's subtitles are not too bad.

J

PS Should not be long until you can like the new software!

Edited By JasonB on 18/03/2021 14:47:55

[Niels AbildgaardParticipant @nielsabildgaard33719]

Posted by mgnbuk on 18/03/2021 13:45:29:

The bed design on these doesn't seem too good if it is similar to the model Nils shows – in one of his other posts Nils showed one of these machines stripped down & I was struck by how poorly the headstock is supported due to the amount of metal removed underneath it to make space for the motor. Most Western lathes continue the bed at it's full form (or beef it up) under the headstock to properly support it.

.

 

The bed of the 180/210 lathes is Lousy but the Canadian gentleman lathe is a 250 and frame is miles better than Myfords and on par with Southbend/Boxfords

 

 

 

 

Edited By Niels Abildgaard on 18/03/2021 15:44:35

[Ex contributorParticipant @mgnbuk]

Thanks for the clarification Nils.

The lowest photo is the one I was thinking of – light use only there, I think !

Nigel B.

[Niels AbildgaardParticipant @nielsabildgaard33719]

Posted by mgnbuk on 18/03/2021 15:48:10:

Thanks for the clarification Nils.

The lowest photo is the one I was thinking of – light use only there, I think !

Nigel B.

 

 

It depends on what You define as ligth

Modified 180 Lathe

 

it was done with a sewing machine motor

Stiffened 180 lathe bed

 

Edited By Niels Abildgaard on 18/03/2021 16:34:12

[Ed PageParticipant @edpage45070]

Quite the response!

I can't address everyone so I guess I will generalise the response, also, I'm originally from the UK. I've owned an old Pooles, Denford Viceroy and Harrison M250. Never had rigidity issues with any of those, but they were good old lathes. Some nice big lathes over here, like 4m long beds or more for cheap, but everything smaller is new Chinese. Going from model lathes to industrial and back to a model lathe is quite a step down. 25HP back to 1HP. Funny that I buy old HSS end mills over here for aluminum, Presto, my home town, take a guess.

I know I'm pushing the lathe a little too much, the original idea was to make one from scratch and have some castings made, wish I had now, but that is a project for another time, but now I don't have access to an industrial machine. Thankfully I will get a decent amount back when it comes to part with this lathe due to the market, plus I have all the original manual parts for it as I'd keep the servos. A little bit of info, 10 inch chuck, 3 inch spindle bore, linear rails, 10 inch swing over slide, 20 inch between centres, gang tooling, 3 or 5HP direct drive, 1500 rpm max. Which brings me to the spindle bore on lathes. If you want a 3 inch bore then you need a 10 inch chuck, that comes with a very long bed, probably over 80 inches. Some CNC's are 36 inches, but the machine is still huge and requires big power. What are peoples thoughts on a hobby / small industrial lathe with a larger spindle bore and shorter bed? I personally hate end pieces, and working at the tail stock if its something long.

So, 1 inch tools are cheap and better quality than the hobby stuff. Round tool holder because it's quicker to make it from round in a lathe rather than mill it, plus round stock is cheaper. Tool post is 3 inch round 120ksi steel. Grooving tool is set length, not adjustable. Insert is 2mm wide, feed 0.04 to 0.12mm/rev, they are intended for light turning. Slowest speed of lathe is 150rpm, but it's belt driven. I will mostly be turning between 1 to 3 inch round, the 5 inch piston was over doing it, but cheaper myself than outsourced, plus it worked out on the mill in the end. I want the lathe mostly for threading, but a 12TPI -1.5 inch thread I still think it's not rigid enough. I agree with the z axis ball screw coupler, I do have to limit the speed because of this. HSS definitely would have been an option, but at 5 inch the surface speed would have been about 200SFM, that's on low carbon steel, in my experience it would not have lasted long, 10 parts total, 2 thou tolerance. HSS is diabolical over here, all comes from China.

I took the plates off the lathe to take a look at the castings, as in the pictures, thankfully it's reasonable unlike the one in Niels post, that is pretty sketchy, can't believe someone would design something like that. I might get a 3HP servo to directly drive the lathe with a timing belt since I will need an encoder for the threading, the motor will then be used on my later CNC project, maybe 5HP depending on the price. Would love to make chips instead of razor wire!

[Ed PageParticipant @edpage45070]

Anyway, only managed a few hours on the lathe, busy day. I took apart the saddle assembly to see if it sat on the bed correctly. Look at the pictures carefully, the vee has contact all the way down, the flat has one tiny contact patch in the corner, looks like this is a big issue. Overall not a lot of contact. I broke out the Sandvik scraper, something I've never done before. I probably took a thou off all surfaces and about 3 thou off the bad corner. Sorry no pictures, my cross-hatching technique didn't look pretty. That's all I managed, but now the slide sticks with oil suction to the bed, I also put force in every direction with a DTI to find nothing moved, I'm happy with that result.

There are two gibs underneath the bed, neither adjustable, but there is a 4 thou gap on both front and back now. I've seen a few recommendations from adjustable gibs to bearings. There isn't much room to play with but I am thinking of something spring loaded, maybe some spring steel strips with a bronze blade? Let me know your thoughts.

I really like the modifications made to the top slides, especially that milled tool post. I have a large piece of steel I'm going to work on in the next day, it will bolt directly to the top slide at its full length. I will then bolt some gang tooling to that, will update in the next 24 hours. It will reduce my swing over the slide to about 6 inches, but I can't imagine ever needing that.

So overall you get what you pay for, and this lathe is better than none at all.

[HopperParticipant @hopper]

Yes when you see inside the headstock, it does not look overly massive. The word Gibraltar certainly does not come to mind.

It's hard to tell from the pics, but it looks like yours has the common shortcoming of many Chinese lathes: The bed width is narrow in comparison to the spindle height. It looks like the bed width is quite a bit less than the spindle height. So the triangle between the bed shears and the spindle is a tall, narrow based one. That's inherently less stable than many better quality industrial lathes and toolroom lathes that have the bed considerably wider than the distance from bed to spindle, so they have a flattish, broad-based triangle to spread cutting forces over. So between the headstock bulk and the bed wiidth, you are never going to get the type of rigidity you are used to in industry.

But as you've seen, going through and "blueprinting" the machine should tighten it up considerably.

I wouild not go with springs under those saddle lift plates. Cutting forces will overcome those springs. Best to do like Myford and most others and machine either the lift plates or the area where they mount on the bottom of the saddle so strips of shim can be used to get the desired half a thou or so clearance there. Anyway, if yoiu are plunging straight in you should have the carriage locked, which ideally should lock on that rear bed way. Might be an addition to consider if it does not have one.

I know you like that 1" tool and the overhang would be fine on a big rigid industrial machine but is just begging for trouble on these smaller flimsier machines.

If you go through as you are and eliminate all these little things, the sum total should put you ahead of where you were. But you are not going to make a Monarch 10EE out it due to the core design and build factors.

[Ed PageParticipant @edpage45070]

Hi Hopper,

Yes the spindle is definitely higher than the width of the bed, plus due to the diameter of the piece I was turning the tool holder was to the front of the bed. I knew it was dodgy, but sometimes you got to do what you got to do. I really wish I had gone for the part-off blades that are adjustable, I should have known better, but now it is too late.

I totally forgot about the bed lock, and I definitely used it on all of my previous lathes. The big issue now is that it is CNC and I'll be mostly using it for threading, so while I could lock it for grooving, I definitely cannot for threading. I have about 100 threads to cut, 0.75 – 16, that will be its next challenge when I sort out the encoder. I then have some 1.5 – 12 threads to do after than. All in 4140, which is 120KSI or thereabouts.

I will see what the others say about the gibs first before I make my final decision, but I do think you might be right in that the springs will just give, shimming would also be fairly easy to do, I just hope the bed is the same thickness all the way down.

I'm probably going to remove the head in the next day or two to see how that's attached, maybe replace the bolts with some studs. Worst case I'll get some nickel rods and weld it together, I might also put a tiny bit more pre-load on the headstock bearings, although haven't checked those to see if they're too tight or too loose yet, but there definitely is no play in it by hand.

[Author]

Posts