Pre-1960 machine tools: the full inspection guide

A machine built in 1955 has had seventy-one years to accumulate wear, and nothing about German engineering exempts it from that. What age changes is where the risk sits. On a five-year-old CNC machine, the question is whether the control is still supported. On a 1955 grinder the control is a handwheel, and the questions are entirely different.

Most of the risk on a pre-1960 machine tool is specific, and most of it is visible during an inspection if you know where to look. This guide covers what to ask before you travel, how to read the iron once you are standing in front of it, what a rebuild actually involves, and the compliance position in the UK and EU, which is more favourable to old machines than most buyers expect.

Before you travel

A wasted trip to a machine four countries away is expensive, and most of them are avoidable with four questions asked in writing.

What tooling is included? Ask for an itemised list, not a reassurance. This is the single biggest swing in what an old machine is worth and it is consistently underestimated. A lathe without its change gears may be unable to cut a thread at all, and having a set cut can cost more than the machine did. Chucks, steadies, faceplates, collets, taper attachments, arbors and toolposts are individually replaceable and collectively worth a substantial fraction of the purchase. When a listing says accessories are included, as the 1957 WMW turret lathe does, get the inventory before you book travel.

Was it under power when it came out? There is a large difference between a machine running in a working shop last month and one that has stood in a building since the 1990s. Good listings say which. The 1943 Wohlenberg lathe is described as dismantled with inspection available, and the WMW as untested. Neither is a problem if the price reflects it and you can recommission. Both are a problem if you assumed otherwise.

Can it be demonstrated cutting? Not spinning. Cutting. A motor that runs tells you the motor runs. A machine taking a cut tells you about rigidity, chatter and finish, which is the thing you are actually buying.

What does it weigh, and what is the supply? Weight determines your rigging bill and whether your floor takes it. Supply voltage determines whether you can switch it on. Both are cheaper to find out now.

How these machines were meant to be tested

There is an established way of measuring whether a machine tool is accurate, and it is roughly as old as the machines themselves. Georg Schlesinger, working in Berlin, set out the first rigorous acceptance tests for machine tools and defined the minimum geometric standards a machine had to meet to be considered fit for production work. His Testing Machine Tools ran to a seventh edition in 1966 and was the reference the industry used for decades. The methods and instruments have moved on, but the principles were carried into the modern ISO 230 series, which still specifies geometric tests for straightness, squareness and spindle runout.

You are not going to run a full Schlesinger test in somebody's yard. But three of its checks are quick, need very little equipment, and will tell you most of what you need to know.

Reading way wear

Wear concentrates where the work happens, which on a lathe means the first stretch of bed in front of the headstock. A bed worn unevenly makes the machine cut a taper that no adjustment will remove.

The useful trick is that the machine carries its own reference. The far end of the bed, past where the saddle normally travels, is usually close to unworn. Measure the ways there with a micrometer, then measure the same section in the heavily used zone near the headstock. The difference between the two is your wear figure, and you did not need the original factory specification to find it.

Then traverse the saddle along the full length of the bed by hand and feel for sections that go tight or slack. You are feeling for the same thing the micrometer told you, in a way that also picks up twist and damage.

The two-collar test

The proof is a cut. Mount a length of bar or tubing in the chuck with two collars several inches apart, take one light pass across both, and measure each collar. If they differ, the machine is not cutting parallel, and the cause is wear, alignment, or both. This is the practical descendant of the turning tests in the acceptance standards, and it is the single most informative five minutes you can spend on a lathe.

Spindle runout

Put a dial indicator on a magnetic base on the bed, bring the stylus against the spindle taper or a test bar held in the spindle, and rotate by hand. The figure you read is total indicator reading; actual eccentricity is half of it.

What counts as acceptable depends entirely on the machine and the work. A precision grinder is held to a different standard than a jobbing lathe. As an indication, a manual lathe with sleeve bearings is generally expected to hold around 0.001in or better at the spindle nose, and three-jaw chuck runout of 0.003in to 0.005in is common and workable.

One caution that catches people: on plain-bearing machines, up to about 0.001in of spindle movement under cutting load is the oil film deflecting, not the bearing being worn out. That is normal behaviour for the design. Do not condemn a good spindle for it.

Backlash

Backlash in the cross slide and leadscrews is normal on conventional machines and operators work around it every day. Measure it, write the number down, and treat it as a cost rather than a verdict. Nuts can usually be remade.

What a rebuild actually involves

"The ways can be reground" is true, and it understates the job considerably.

Regrinding means stripping the machine to the bare bed casting and setting it up on a bed grinder, which is often itself an old planer converted for the purpose. Once the ways are ground, everything that references them has moved. The saddle has to be rebuilt to match the new surfaces, usually with Turcite or Rulon bonded on and machined in. The headstock and tailstock then have to be realigned to the corrected height. Finish work is hand scraping, where a good result means eight to ten bearing points per square inch across the surface.

This is specialist work measured in weeks, not days. Indicative figures from rebuilders put bed grinding in the region of 100 US dollars per foot, a four-foot bed regrind with carriage work at somewhere over 1,000 US dollars, and a full medium lathe rebuild in the region of 4,000 US dollars. Treat those as orders of magnitude rather than quotes: the only number that matters is the one your rebuilder gives you for your machine.

The reason to know the shape of this work is that it tells you when to walk away. A machine needing a full rebuild plus new tooling plus a rigging bill can quietly cost more than a sound machine from the 1990s that does the same job.

The machine that has been standing

Every surviving pre-1960 machine took one of two routes to get here. It either earned its place on a shop floor for decades, or it stopped being needed and stood in a corner. Both leave you a machine that still exists. They are not the same purchase.

On a machine that has stood, look for what moisture has done. Surface rust on a bed is cosmetic and cleans up. Rust at a bearing seal is not: once the seal is compromised, air and moisture reach the bearing, the lubricant dries out, and the bearing is on its way to seizing whether or not anything looks wrong from outside.

Check that everything moves before you assume it does. Seized ways, dried grease in gearboxes and a spindle that will not turn by hand are all normal on a machine that has stood for twenty years, and all of them are work. If somebody has already attempted a clean-up, ask how: abrasive grit that finds its way into bearings and gears does more harm than the rust did.

One thing is not recoverable. A part that has corroded through a load-bearing section does not go back into service however good it looks once it has been cleaned and painted.

The additions are usually the risk, not the iron

This is the part that surprises buyers. Cast iron does not fatigue standing still, and a 1955 bed that was scraped flat is still flat. What tends to fail is whatever was fitted later.

The 1953 Waldrich Siegen portal mill is listed as CNC, meaning a control retrofitted decades after it was built. A control system from the 1980s is now considerably harder to support than the 1953 casting it is bolted to, because it is electronic, proprietary, and made in far smaller numbers than the castings ever were. The same logic applies to a rewired motor starter, a fitted digital readout, or a replaced drive. When you assess one of these machines, price the iron and the electronics as two separate risks.

Mechanical spares are usually a smaller problem than people fear. Machine tool bearings have long used standard envelope dimensions, so substitution is often possible even when the original part number is long gone, and specialist suppliers will source or manufacture precision spindle bearings to order. Where a part is genuinely bespoke, machine dismantlers who break old machines for spares are a real route, and enthusiast communities maintain cross-reference lists of substitute bearings for exactly this purpose. Electronics are the harder problem, which is the opposite of what most buyers assume.

Power supply

These machines expect three-phase. In most European industrial premises that is not an issue, and you can skip this section. If you are putting a machine into a small workshop on single-phase supply, it is the first thing to solve.

There is a specific trap worth knowing: a variable frequency drive taking 230V single-phase input will not produce 400V three-phase output unless it is a high-voltage series drive. European machines are generally 400V. Running a 400V motor from a 230V-input drive without a step-up transformer gives you poor performance, nuisance trips, or a damaged motor. Check the machine's nameplate voltage against what the drive can actually deliver.

Beyond that, the choice is between a VFD and a rotary phase converter. A VFD is usually the cheaper and more compact answer for a single machine under about 3HP, and adds speed control and soft start. A rotary phase converter generates a genuine third leg using an idler motor, gives you full nameplate power, will run several machines, and leaves the machine's original contactors and controls working as designed, which matters on a machine whose control gear is part of what you bought.

Asbestos: where it actually is

Asbestos in machinery of this era is not mainly a wiring issue, though insulated wire loom is one place it appears. The more common locations are friction and sealing materials: clutch facings, brake linings, gaskets, packing, heat shields and some bearing housings.

That matters because of what people do to old machines. Undisturbed asbestos in a clutch pack is not releasing fibres. The same clutch pack being stripped, scraped, ground or blown out with an airline is. Identify it before any work starts, not after, and handle it under the rules that apply where you are.

Presses, shears and anything with a clutch

Mechanical presses and shears deserve their own paragraph because the failure mode is specific and severe. The classic mechanical press accident is the unintended repeat stroke, where the ram fails to stop at the top of its travel and cycles again, and the usual causes are a worn clutch or brake or a fault in the control circuit.

Modern practice requires an anti-repeat feature and a single-stroke mechanism in the control system, with the clutch and brake condition checked on a regular schedule. A press or shear built in the 1950s, such as this 1955 Schuler sheet metal shear, predates all of that. Assume the guarding and controls need bringing up to current standards, budget for it, and understand that this is not optional if anybody is going to be employed to operate it.

The legal position, which is better than people assume

Buyers often worry that a 1955 machine with no CE mark is a compliance problem. Usually it is not, and the reason is a date.

Machinery first supplied into the European Community before 1 January 1995 was never required to carry a CE mark, and it does not acquire that requirement by being repaired or refurbished to its original specification. A 1955 grinder that has been in Germany since 1955 legitimately has no CE mark, and its absence tells you nothing bad.

Three things change that position, and they are worth knowing before you buy:

• Importing from outside the EU. Crossing the border counts as first supply into the market. A pre-1960 machine brought in from outside the EU is treated as being placed on the market for the first time, and the compliance obligations follow it in.
• Substantial modification. If a machine's specification has been enhanced or changed, it may be treated as a new machine, with the full set of rules applying. A CNC retrofit on a 1953 mill is exactly the kind of change that raises this question.
• The 2027 changeover. Machinery placed on the EU market from 20 January 2027 falls under Regulation (EU) 2023/1230 rather than Directive 2006/42/EC. If your purchase straddles that date, check which applies.

What does not change is your duty as an employer. In the UK, the Provision and Use of Work Equipment Regulations apply to all work equipment regardless of age, and to second-hand, hired and salvaged machines exactly as they apply to new ones. Equipment must be suitable, safe, maintained and inspected, and used by people who have been trained on it. "It was working when I bought it" is not a defence, and "sold as seen" transfers nothing: if the machine arrives without guards, fitting guards before anyone uses it is your obligation. For pre-1995 machines the practical route is a risk assessment against the essential safety requirements, and then doing whatever that assessment says.

None of this is a reason not to buy. It is a reason to know which of those four situations you are in before you agree a price, and to take proper advice where the answer is not obvious.

The cost that is not the machine

On equipment at this weight, moving it is a major line item and on a cheap machine it can exceed the purchase price.

The drivers are weight, access at both ends, how much disassembly is needed, distance, and whether the load is oversized enough to need permits and escorts. The 1940 Lindner drill is around 2,500kg, which is straightforward for a rigger with the right kit. The Waldrich Siegen portal mill, with 14 metres of table travel, is a different proposition entirely: specialist lifting, probable foundation work, and levelling on arrival before it will hold any of the accuracy you paid for.

Riggers generally price on crew hours plus equipment, and the variables above move the total far more than distance alone does. Get a firm quote against the actual machine and both actual sites before you agree a price, not after. It is the most commonly underestimated number in the whole purchase.

So is it worth buying?

There are two buyers here and the honest answer differs for each.

If you have maintenance capability in-house or a working relationship with a machine rebuilder, and what you need is capability rather than guaranteed uptime, this equipment is frequently very good value. The castings carry more mass than modern equivalents, the mechanics can be diagnosed without proprietary tooling or a service contract, and you are not dependent on a manufacturer that in several cases stopped existing decades ago.

If you need a machine that works on the morning it arrives, with support behind it and no rebuild budget, a machine from the 1950s is the wrong purchase no matter how well it was built. Something from the 1990s will serve you better.

On price, most machines of this age are listed on request rather than with a figure attached, and that is not evasion: condition, tooling and location move the number so far that a published price would be close to meaningless. We have written separately about how that works and how to get a real number out of it.

What the wear tells you

Everything above comes down to one question, and the machine answers it itself.

A bed worn evenly along its length was used properly for decades by people who moved their work around. A bed with a sharp step of wear in the first foot was run hard on repetitive short work. A machine with almost no wear and significant rust did very little, and then sat. None of those three is disqualifying, but they are three different machines at three different prices, and no listing is going to tell you which one you are looking at.

The iron keeps its own record. An inspection is just reading it, and the ability to read it is what separates a good buy at this age from an expensive one.

For what these machines were built to do in the first place, see the companion piece on the oldest machines listed on Machine Octopus. For buying used machinery generally, at any age, the 2026 buying checklist covers the ground that applies across the board.