A KRV2000 Called Matilda

This is my first post on the forum, but have been lurking for some time gathering information and knowhow, so I thought I’d share our (mis)adventures with our current project(s).

We shut down our engineering business 3 years ago following my partners illness and we both now want to dabble with machine tools again, so are keen to CNC retrofit a milling machine for disabled use.

Earlier this year we purchased a new (to us) 1996 KRV2000 mill to replace our RF45 clone (Chester Superlux) we had from new in 2012. The RF45 clone was great, but after much consideration we chose not to go with converting the RF45 because it just didn’t have enough travel and was not rigid enough for what we wanted. We did consider a “small” bed mill like a SMX2500, but it would have needed to be disassembled to get it in the new workshop, and would have been too big anyway.

The plan is to do a CNC conversion on it with linuxcnc 7i76e/7i85 combination controller. This will enable dual loop feedback with feedback loop to the controller coming from some linear scales. The machine was used as a manual machine, and a few additions made like some slideways covers, spindle halo light and a digital quill readout. We also cut and rewelded the chip tray to enable easier access around the machine and reduce it’s footprint, along with rear splashguards on the column to contain some of the chips.

We decided to concentrate on servicing the head first to keep the machine in service more easily. Apart from new brake shoes and a single spindle bearing it was in pretty good shape.

We have named her Matilda and she was filthy. We were really shocked at just how much grime and disgusting fudge was on her that in a backlash to this we decided on a ‘force you to clean it’ white paint job! It was almost as if it had never been cleaned from new.

Pics and details to follow as we complete the rebuild and conversion.

Following the relatively smooth experience with the head, and getting her back up and running to machine some parts, we turned attention to the lower half of the machine. The sump was the area we really were not looking forward to which was disgusting and stank as soon as we started to investigate. Unfortunately, the only access point to the sump is via a door half way up the column, unlike some more sensible designs which have the door lower down on the back for easier access to the coolant reservoir. A hex socket grub screw on the foot was also installed on the machine. However we soon discovered the hole is not located at the floor level of the sump.

After a bit of a battle, we managed to remove the coolant pump with the idea of keeping it for later use (after cleaning!). A roll of paper rags, coupled with PPE that would make a bioweapon lab safety officer proud, and a strong constitution held the day. This was truly disgusting not to mention a serious health hazard – it got everywhere – even up the wall of the workshop. The smell still lingered after a couple of days, so we decided to order a couple of kilos of small crystal sodium hydroxide to try and saponify anything remaining in the sump and flush out any debris and undissolved “distgusto”. Not pellets which we feared could be problematic due to localised exotherm at the bottom of a plastic container when mixing. The solution was mixed in a plastic jug and was agitated as soon as the crystals went in to allow them to quickly dissipate any heat. An irrigation system was constructed using the drain hole on the foot coupled onto a threaded hose connector (originally from the milling table) through a somewhat elaborate guttering system channelled outside to the storm drain. The hose from the connector was clamped off and a lab wash bottle that we usually use for coolant was filled with the sodium hydroxide solution so we could clean inside the knee at the same time. We made sure to avoid the bearings and bevel gears inside the knee which we covered over as well as the slideways, oil line etc. Wearing PPE and a face shield (sodium hydroxide solution and eyes do not mix without dire consequences), we spent about an hour repeatedly irrigating the inside of the knee in several bursts to allow time for any reaction to take place until the sump filled up and overflowed onto the foot. Areas of soapy foaming confirmed that reaction was occurring. Then it was left for about an hour to soak. The hose was unclamped and the solution started to drain until there was about 10mm left in the bottom which couldn’t drain due to the exit hole height. The wash bottle was filled with water and the inside of the knee was carefully rinsed and dried. A hose pipe was used to flush the sump and dilute the sodium hydroxide solution. It was quite successful if very hazardous. It also seemed to selectively dissolve one of the layers of paint!

A couple of days later the mill was clean, but there was still a slight lingering smell, and having been through what we regarded as a horror show that we would rather forget, we reluctantly decided to do the whole thing over again but this time with a disinfectant. We used Jeyes fluid in the strongest recommended concentration and using the same techniques as earlier but leaving the knee alone. This did the trick – smell vanquished. Well almost – its been replaced by the slowly fading odour of the disinfectant. But at least its clean.

Following this gruelling battle, we have decided to seal the sump and use a remote coolant system using the original pump, a newly fabricated steel reservoir and a vfd to give us more control over the flow rate. We did something similar for our lathe that has a single phase pump where we used an SCR rather than a vfd. We aren’t fans of bathing in coolant!

I ought to add a warning.

Sodium Hydroxide solution can be dangerous, what is described above should not to be attempted, especially in the way that we did it, without a full risk assessment and full awareness of the hazards.

Now that we sealed the sump, there was quite a lot of unused space inside the column. So we decided to fabricate an insert in the form of an enclosure for the 2 vfds. This would ensure that the portal to hell was well and truly closed forever (unless we needed to change the wiring) and ensure minimal flashbacks to past traumas. It did take a little reshaping of the “portal” to fit the enclosure due to the sharp corners of the welded box.

While we were disassembling and cleaning the oiler system we had a slight mishap with the aluminium tubing to one side of the saddle where the oil was delivered to the knee/saddle slideway. So one of the lines was replaced with nylon tubing. All lines were blown through with compressed air and checked to make sure oil would be delivered through them.

Next, we turned our attention to the paintwork, which after 25 years of knocks, chips, coolant and oil (and strong alkali solution!), was in a sorry state. The column and knee had been coated with several layers of red oxide, thick filler primer and paint which had been penetrated by oil and grime. The interlayer adhesion was poor, especially the filler primer to the underlying red oxide. Since the column was not too bad it was decided to just wet sand with 240 grit to provide a key for the new paint, but the knee was scraped with a chisel until all the paint and filler primer was removed leaving the red oxide first layer. This was then wet sanded with the same 240 grit and then cleaned and painted with 2 layers of white paint. We used the ArcForce fast dry machinery enamel which dries really fast and gives a reasonably good finish with a brush or roller.

We had read reports on this forum and elsewhere that this design of mill suffers from a “springy lash” reportedly due to the yoke design. We had got a little annoyed with how the DRO would register a “creep” moments after setting a position, especially on the Y axis and it wasn’t something that we had encountered on our previous mill. This was despite spending some time playing with the gib strips to get them tuned in as well as possible. Before re-assembly, the yoke was measured and a new one based on the Hiwin and Elrod ones was designed with a view to eliminating this problem. All the measurements on this mill seem to be a combination of imperial and metric with imperial fasteners and metric dimensions for some (not all) of the castings. The yoke had very different measurements to what we were expecting from the Hiwin drawings for a Bridgeport conversion. Also a concern, was the clearance issue of the Y axis nut mount part of the yoke with the knee lift mechanism. The leadscrew nuts are metric with a 40mm OD in a dual nut backlash compensation configuration. As expected even before we inspected the yoke, we weren’t happy with just boring out the original one, or compromising on the new ballscrew to a smaller diameter to make it fit.

The next issue was where to get a chunk of material to make this new yoke. We did consider using cast iron from an old anvil we had lying around, but we couldn’t really justify hacking it up. After quite a bit of searching, we found a recommendation for a niche supplier of materials in Darlington that had lots of interesting alloys and polymer materials, and specialised in small quantities for DIY. Thanks M-Machine! We ordered a rather large chunk of steel that would be just the job, and a few other bits to do a Z axis yoke (more on that later).
We decided on some Gten 3205 C5 ballscrews and FDID double ballnuts which were smaller than the usual FDUD to ensure clearance with the knee mechanism, all from Motion Control Products (formerly Zapp Automation, I think, who we have used many times in the past for other projects).

Another surprise we discovered was the use of standard deep groove bearings on the X and Y axes for the original leadscrews. Reading through the installation notes for the ballscrews, it recommends angular contact bearings at the drive end with preferably a floating support at the other end. We found a Dutch supplier of cnc parts selling C3 FK25 and FF25 bearings for a very reasonable price. Thanks to Damencnc for getting them here on Plague Island within days of ordering and before the Kent lorry debacle during Christmas 2020.

So we spent the last few days finishing cleaning and inspecting the table, and reassembling the mill. Took a bit of playing around with the backlash compensation setting of the leadscrew nuts. We found that there was quite a bit of wear in the central part of the leadscrew which made setting the leadscrew nuts a compromise between being really tight at the ends of travel with backlash in the central area at a minimum, and being just too loose in the middle. After what felt like hours of fiddling, we then embarked on the tricky manoeuvre of sliding the table back on and adjusting the gib. She lives again!

The belts and pulleys turned up, enabling some final fine tuning of the design of the drive end plates of each axis to optimise centre distance between the drive and ballscrew pulleys. The ballscrews also arrived a few days later, so some end machining will have to be done.

Hi Chris - Matilda will be an excellent machine. Everything needs TLC. Great story.. Keep at it Peter

Hi Peter, thanks for the encouragement.

Just seen your Milli thread, all I can say is WOW! We did look at composites before we got Matilda when we were weighing up the options, but we were nowhere near confident enough to embark on a project with composites, especially for metalworking. I'm not well enough versed in FEA to attempt to design something that would have been guaranteed to work for us and do what we wanted. Perhaps when we upgrade our little chinese router we might have a go at something small! Following your thread with amazement and great interest.

Chrissy

So…..the next step was to start machining the endplates for the X and Y axes. We had some 15mm aluminium plate hanging around from an aborted CNC router upgrade project that we were going to do but never got around to it. Luck would have it that there were 2 plates 450mm x 150mm which was just enough to do the job. We decided that we would use jockey wheels to tension the belts as it enables the motor position to be fixed and makes fine tuning easier. This would be done after the main machining so that we could mock it up with the belts and pulleys to get the optimum positioning.

We decided on having the X axis motor at the front on the right hand side just below the surface of the table. This way we avoid having the thing stuck out down low like the commercial ones do where they give the machine a larger overall footprint, and it can be covered over with a front cover that extends over the limit switches as well. It also gives an extra area for the inevitable collet or spanner etc that always needs to be put down but still be available to hand! The use of the FF25/FK25 bearings meant that we could do fine adjustment of the position of the ballscrew, just in case we made any errors in the measuring of the exact leadscrew position in relation to the mounting holes at the end of the table, and also as we are machining a brand new yoke. We checked it several times, but sod’s law can strike at any time. In the end we found that we need not have worried.

The Y axis endplate has been designed to be quite short and enable the motor to be kept within the knee like several commercial designs. We didn’t like the extended plates that drop either to the side at an angle or below the knee casting as these leave the motor exposed to chips, slideways oil and coolant making cleaning down the machine more of a chore. The only drawback is that we needed to cut a hole in the knee to do this, and we are always a sucker for making things more interesting!

We are really looking forward to not having to use that bl**dy boring head again once we have the conversion done. It has been the source of much annoyance – it just refuses to be consistent. The rotary table is not very popular either, being a bit too heavy for us to lift without being concerned we might get hurt. It already caused one injury, and it will take more than a Chloe Ting workout programme to get enough strength to lift it with confidence.