Most Chinese aluminium sliding door factories carry some version of the same sentence: German technology. It sits next to the photographs of the CNC centres and the export map. It is rarely a lie, and it is never a certificate. The problem is not that the claim is false. The problem is that four different claims share one phrase, and the one buyers check is the one that matters least for a sliding door.
This article takes the phrase apart. It is written from the factory side, and it deliberately does the thing these pages usually avoid: it tells you which claims we can put in writing, which claims only the shop floor can settle, and which claim a buyer should never assume from an adjective.
When a factory says German technology, it is usually asserting one or more of four things. They are not equivalent, they are not equally easy to verify, and they do not affect your door equally.
| Claim | What it actually asserts | How you check it | What it decides on your door |
|---|---|---|---|
| German-brand components | Named hardware brands are fitted: transmission boxes, multi-point locks, handles, rollers, side-pressure gear | Brand and series named on the purchase order, supply voucher, batch number | How the door feels every day, and how long the mechanism lasts |
| German-made machinery | The strip insertion and crimping line, the cutting and milling centres, sometimes the coating line | Stand in front of it. Read the nameplate. Ask for the process records it produces | Whether panel 400 is square like the sample, and whether the thermal break holds |
| Designed and tested to German or European standards | Composite profile shear to EN 14024, whole-door classification to EN 14351-1 with EN 12207, EN 12208 and EN 12210, Euro-groove hardware interface | Test report naming the model, the laboratory, the date and the test conditions | What you can submit to a building authority or a consultant |
| German ownership, or made in Germany | A German legal entity stands behind the product | Business registration, the legal name that issues the invoice | Who you can sue, and under which law |
The first claim is the one every buyer checks, because it is the only one you can check with your hands. Turn the handle, read the logo, feel the weight. That is real, and it matters.
But notice where a sliding door actually fails. It does not fail at the handle. It fails when the panel stops gliding, when the corner weeps in a storm, when the roller sits in water it was never meant to sit in, when the two halves of a thermal break creep apart over a decade and take the frame out of square with them. None of those are hardware failures. All of them are outputs of the second and third claims, and neither can be read off a sample.
A thermally broken aluminium profile is three parts pretending to be one. Two aluminium halves, a polyamide strip between them, and a mechanical bond that has to transfer load from one half to the other across a plastic bridge.
The bond is made in two steps that run on one machine. First a knurling head cuts teeth into the aluminium groove. Then the strip is fed in and sets of hardened wheels roll the aluminium onto it, closing the groove over the strip so the teeth bite. The equipment suppliers who build these lines describe the result in one sentence: the assembled aluminium and strip profiles then act as a composite to transfer loads.
Here is the part that makes this a machine question rather than a worker question. One of the European builders of thermal break assembly lines, the Belgian company Aluro, describes the problem in its own technical writing: a modern profile asks the assembly line for two things at once, a mechanical bond that holds and a finished profile that stays inside tolerance, and the two work against each other. That tension is where the line either earns its place or does not.
Crimp too hard and you get beautiful shear numbers on a profile that is no longer straight. On a sliding door that means the panel does not sit square in the frame, the rollers load unevenly, and the door drags on one side from the day it is installed. Crimp too light and the profile stays perfectly in tolerance while the bond quietly loses capacity, and the failure arrives years later as a slow drift rather than a break.
This is why the machine's calibration is the product. It is also why a buyer should never accept a thermal break claim that rests on the material name alone. PA66GF25 is the right material. What it becomes depends on how it was put in.
The standard that governs this is EN 14024, Metal profiles with thermal barrier, mechanical performance, requirements, proof and tests for assessment. The current edition is EN 14024:2023. It is worth knowing three things about it before you ask for a figure.
One useful detail for anyone comparing quotations: the standard publishes creep coefficients for known materials rather than making every buyer test them, and it names PA66 GF25 of a stated minimum density as one of them. If a supplier cannot tell you whether their shear figure is an as-delivered value or an aged value, that is the answer.
We publish a shear resistance of 5,000 N on the AW90 passive window. Two honest qualifications come with that number. It is a published test result for that series, not for the sliding door series, and it is not a project qualification. If you are buying a sliding door, ask for the same figure for the series on your order, tested to a named standard, at a named temperature category, before and after ageing.
A sliding door is made in a fixed sequence. Each station sets something the next station cannot correct. This table is the one to carry onto the floor.
| Station | What it sets | What failure looks like later | What to ask for |
|---|---|---|---|
| Strip insertion, knurling, crimping | Composite action between the two aluminium halves | Frame drifts out of square, panel drags, rollers load unevenly | Shear value for this series, standard named, temperature category named, aged and unaged |
| Corner joining | Frame squareness and the water path at the most vulnerable joint | Water at the corner, diagonal drift, gasket pulled out of the corner | Which method, and corner joint strength data. Our C113 runs a full glue injection process with integrated glued corner brackets; the C120 seals frame, sash and mesh with injection-moulded structural adhesive plus high-strength corner reinforcement and complete corner sealing; the C116 uses triple sealing strips with seamless welded corners |
| Cutting and milling | Lock point alignment, drainage slot position, interlock fit | Lock points that do not all engage; drain slots that miss the chamber they were drawn to drain | Stated cutting tolerance, tool calibration dates, and whether unit four hundred is held to the sample |
| Threshold and drainage forming | Where water goes after it gets past the outer seal, which it will | Water over the inner rail, rollers standing in water, staining at the sill | The drainage path drawn, and which rail height you are getting. The SL112/167 has a 45 mm high-ｄｒｏｐ bottom track with an optional 85 mm water barrier; the SL126/182 runs a 43 mm high track or an 8 mm low track with a patented eave-style water defence and a water-blocking edge on the bottom track; the SL155 uses a passive check valve to speed drainage and stop backflow |
| Glazing and sealed unit assembly | Glass weight, seal life, gas retention | Condensation inside the unit; a panel heavier than the rollers were rated for | The exact glass build in writing, and the roller rating that goes with it. Our SL112/167 is specified with 24A insulated glass; the SL126/182 in the 75 mm variant carries 27A insulating glass with inert gas filling and a butyl rubber gasket at the frame and sash joints |
| Surface treatment | Coating film build, pretreatment quality, colour stability | Chalking, fading, filiform corrosion under a coating defect | Coating specification and film thickness, plus salt spray data for the series you are buying. We publish 240 hours of salt spray resistance on the AW90 |
Two notes on that table. First, the glass line and the roller line have to be read together. Maximum panel sizes are published for an assumed glass build; if you upgrade the glass, the maximum panel comes down. Specify glass weight, roller rating and maximum panel on the same line of the order, or you will get three answers that do not add up.
Second, the coating station is the one nobody photographs, and it is where a finish decision becomes a maintenance decision. A white sliding door and a dark one on the same elevation do not age at the same rate, because dark films absorb more heat and aluminium moves roughly twice as much as steel for the same temperature change. Ask what pretreatment the line runs and how often the bath chemistry is checked, not just which powder brand is in the hopper.
Most sliding doors can be adjusted after installation. Rollers have adjustment, tracks can be shimmed, and a panel that is slightly out can usually be brought back. There is one type where none of that is true.
A pocket sliding door disappears into a wall pocket. Once the wall is closed and finished, you cannot get at the track, the rollers or the head guide without opening the wall. That single fact changes what you should demand from the factory: not a better door, but a documented one. Cutting tolerance, corner squareness, roller rating and the drainage path all have to be right the first time, because there is no second time.
The same logic applies, less severely, to a bypass sliding door. Two panels pass each other on parallel tracks, so the interlock stiles are the most visible element on the elevation and the least forgiving of a milling error. If those two stiles are not coplanar, you will see it every time you look at the door, and no amount of adjustment at the rollers will hide it.
If you are going to visit, do not spend the time in the showroom. One Chinese factory that publishes its own customer audit checklist makes the point plainly: a supplier who will only show you the showroom is telling you something, and it is not good. Walk the line in production order and ask one question at each station.
Two more signals worth reading while you walk. In a well-run plant, rejected parts are put where you can see them, not hidden behind a curtain, and each assembly carries a card recording its source, the operator and the machine, so a defect can be traced to a cause. And look at the production boards: a digital dashboard whose figures have not moved in three months is decoration, not management.
We are not going to claim a machine roster we have not published. What we can do is separate what is on the record from what has to be requested, because a quotation that blurs those two columns is where most disputes start.
| On the record | You have to ask |
|---|---|
| 6060-T66 alloy across every series | The machine list. We do not publish one, so ask on the call and ask what each line does |
| Wall thickness by series: 1.8 to 2.0 mm on the casement and tilt-and-turn series, 2.0 mm with local reinforcement to 3.0 mm on the SL126/182, 2.2 to 2.5 mm on the SL155 | Whether the stated thickness is nominal or minimum, in writing |
| PA66GF25 thermal break material; 34 to 44 mm multi-chamber strips on the C116 | Shear value for the specific series on your order, with the standard, temperature category and ageing condition named. Our published 5,000 N figure is for the AW90 |
| Thermal values K 1.3 on the AW90, 1.6 on the C120, 1.7 on the C116 | K or U value for the sliding door series. These are not published, and the figures above are system values rather than whole-door values. Solar heat gain is not published for any series |
| 240-hour salt spray resistance and 5,000 N shear on the AW90 | The same two figures for the sliding series you are buying, and for your coating choice. These are published test results, not a coastal qualification, and they do not transfer between series |
| Hardware by series: Winkhaus on the AW90, ROTO transmission boxes and anti-pry lock points on the C120, German ROTO side-pressure hardware on the TS130, CMECH on the C113 tested to 30,000 cycles, Wehag available as an upgrade on the C116 | Roller cycle testing under load for the sliding series. The published 30,000 cycles is a hardware bench test on the C113, not a service life and not a figure for the door rollers |
| Roller ratings: 200 kg on the TS130, 220 kg on the SL112/167, 400 kg on the twenty-wheel SL155 | The panel dimension that rating was achieved at, and what it becomes if you specify heavier glass |
| Maximum single sash: 2800 by 2600 mm on the SL155, 2.5 by 3.0 m on the SL126/182 | Classification numbers for the complete door assembly under the standard your authority uses |
| Two production centres in Foshan and Zhaoqing, 100,000 square metres | Which centre your order will actually be built in, and the current lead time for that line |
This is worth saying plainly, because a factory page that pretends otherwise is not a page you should trust.
A European system house usually holds a complete, third-party classification for the whole assembly, under a named standard, with factory surveillance behind it, and publishes it as a system rather than as a series. That document is what a consultant or a building authority can accept with the least argument. A factory buying the same hardware brands has to assemble that evidence per order, and the published values often cover a series or a component rather than the finished door.
Two related points follow from that, and both are frequently misunderstood. A quality management certificate such as ISO 9001 says a documented quality system exists; it does not independently certify the structural, thermal or weather performance of a window. And a CE marking is not a universal passport: it addresses the European framework it was issued under, not the suitability of the product for every country or project.
For a small project, a single dwelling, or a job where the local authority will not ask for classified data, the difference may not be worth paying for. For a development with a specification, a facade consultant and a handover file, it very often is. That is a decision about your project, not about our pride.
These convert the phrase from an adjective into a line item. Send them as a numbered list and ask for written answers.
| Ask | A good answer | A red flag |
|---|---|---|
| Which machine inserts and crimps the strip on my series, and where is the shear tester | A named line, a tester on the floor, and last month's results tied to jobs | A photograph of a machine with no records behind it |
| What is the shear value for this series, to which standard, at which temperature category | A figure with the standard edition, the category and the laboratory named | We use PA66GF25, with no number attached |
| Aged and unaged values | Both, with the ageing method named | One number, no condition |
| Is the stated wall thickness nominal or minimum | Minimum, stated in writing on the order | The word thickness with no qualifier |
| What tolerance do you hold on the four hundredth unit | A figure, plus a calibration date on the tools | Our workers are very experienced |
| How are the corners joined, and what is the corner strength | A method and test data for it | We use the strongest method available |
| Draw the drainage path on my threshold option | A section drawing with the exit point marked | It drains automatically |
| Give me glass build, roller rating and maximum panel on one line | Three figures that are consistent with each other | Maximum size quoted with no glass build behind it |
| Which hardware brand and series, named on the order | Brand, series and a supply voucher with a batch number | German hardware, or equivalent |
| What can I submit, and to which standard was it tested | A named standard, a named laboratory, and the limits of what the report covers | A certificate with no standard number and no model name |
Almost never, and this is the claim you should check first because it is the easiest to check. The legal name that issues your invoice and appears on the business registration tells you the answer in one step. Everything else on the page can be true while this one is not.
Yes, for a sliding door more than for any other product type. Hardware governs how the door feels to operate and how long the mechanism lasts. It does not govern whether the frame stays square, whether the thermal break keeps its bond, or whether water leaves the track. Those come from the profile assembly and the machining, and they are the failures that arrive after the handle still feels fine.
A shear test result for the exact series on your order, to a named standard edition, at a named temperature category, before and after ageing. It is the one document that can only exist if the factory actually tests what it makes, and it is the one most quotations quietly omit.
Partly. A walk through the line on a live call, with you choosing where the camera stops, will settle whether the strip insertion line and the shear tester exist and are running. It will not settle tolerance on unit four hundred or the state of the coating baths. For a first order, a sample approval and a pre-shipment inspection by a third party usually cost less than correcting failures after installation.
No, and this is where most specification mistakes happen. Our 240-hour salt spray figure and our 5,000 N shear figure are published for the AW90. Our thermal values are published for the AW90, the C120 and the C116, and they are system values rather than whole-door values. A sliding door series has its own numbers, and if we have not published them you should ask for them rather than borrow a number from another series.
No. A larger panel moves the load onto the rollers, the glass and the frame all at once, and the published maximum applies to an assumed glass build. Our SL155 carries 400 kg on a twenty-wheel self-balancing pulley up to 2800 by 2600 mm, and the SL126/182 reaches 2.5 by 3.0 m. Both figures assume a specific build. Upgrade the glass and the maximum panel comes down, which is why glass, roller rating and panel size belong on the same line of the order.
The phrase German technology is not the enemy. Vagueness is. Ask which claim is being made, ask for the one document that proves it, and the phrase stops being marketing and becomes a specification.
