Get the part right before discussing the test items
The first trap in walking aid testing is often not in the laboratory but on the application form. The ISO 11199 series addresses walking aids on which the user supports body weight through both arms, and the series divides internally according to what provides the support and how the device moves. Get the part wrong and the whole test plan has to be rebuilt.
- ISO 11199-1 covers walking frames, where the user grips the handles and advances by lifting or alternately shifting the whole frame, with ferrules in contact with the ground.
- ISO 11199-2 covers wheeled walking frames, known throughout the trade as rollators: wheeled, braked, usually foldable, advanced by pushing.
- ISO 11199-3 covers walking tables, identified by a forearm or upper-limb support platform, so that body weight passes through the forearms and elbows rather than through the palms.
The decision always lands on the same question: through which interface does body weight enter the frame. If it passes through palms and wrists pressing on handles, the product is still ISO 11199-2. Only when it lands on a support platform through the forearms and elbows does it become ISO 11199-3. Having wheels, having brakes and being foldable do not by themselves decide the part, since a walking table may also have wheels and brakes.
| Product form | Main load-bearing interface | Means of movement | Applicable part |
|---|---|---|---|
| Non-wheeled walking frame | Handles, palms | Lifted or alternately shifted | ISO 11199-1 |
| Two- or four-wheeled rollator, including versions with a rest seat | Handles, palms | Pushed, with brakes | ISO 11199-2 |
| Walking table with forearm or upper-limb support platform, may have wheels and brakes | Forearms, elbows | Pushed | ISO 11199-3 |
| Single-arm walking sticks such as elbow crutches | One upper limb, handgrip plus forearm cuff | Single-arm support while stepping | Outside the ISO 11199 series |
Three ways the application typically goes off course, all correctable with the interface criterion above.
A four-wheeled rollator with a rest seat declared as a walking table. The seat is an added rest function. While walking, the user's weight still passes through the handles, so the basic use pattern of "handles carry load, wheels provide movement" is unchanged and the product remains under ISO 11199-2. We have seen "it has a seat, so it must be a walking table" applications more than once, and the cost of replanning is much higher than the cost of one question beforehand.
Treating a tray or a shelf as a forearm support platform. Trays and shelves are for carrying items. They are not designed to bear body weight and come with no matching elbow support or grip structure, so their presence does not change the part. They do, however, need to appear in the declared configuration list, because loading them shifts the centre of gravity of the whole device and the conclusions on stability-related items move with it.
One frame, two forms, only one declared. The same frame fitted with a handle assembly is a wheeled walking frame; fitted with a forearm support assembly it becomes a walking table. Declare only one form and sales of the other are not covered by any report. Confirm the applicable part for each form and issue separate reports.
One more boundary to draw. Single-arm walking sticks, such as elbow crutches and multi-leg canes, where the user supports weight through one upper limb with the grip point ahead of and beside the body, are outside the ISO 11199 series and are not discussed here. If your factory makes both categories, do not expect one report to cover both; which standard applies to that category and which edition to run is confirmed against the current valid version of the standard text and the actual product construction. The mapping of parts to standards can also be cross-checked on the standards index page.
Three things a rollator has that a walking frame does not
What follows is engineering analysis based on load paths, not a statistical statement about failure rates.
The load path in a walking frame is comparatively clean: handle to upright, upright to cross member, cross member to ferrule, ferrule to ground, with several contact points on the ground at once, close to a statically determinate structure. Put wheels on it and at least three things change.
The ground constraint changes. A ferrule is a surface contact with frictional locking; a wheel is a point contact that rolls, and the lateral constraint all but disappears. When the user pushes sideways, the load is no longer shared by ground friction but carried much more by the torsional stiffness of the frame itself and by the wheel forks. This is why, with the same tube diameter and the same wall thickness, removing the ferrules and fitting wheels noticeably reduces the lateral stiffness of the whole device.
Swivel castors add a moment. Front wheels are usually swivel castors whose axle is offset from the upright axis. During steering, or when meeting a lateral obstruction, that offset generates an additional moment about the axis at the root of the upright. Under static loading the component is inconspicuous, but under cyclic loading it is one of the main sources of cracking at the upright-to-base weld and at clamp locations.
Folding and adjustment cut continuous tubes into segments with clearance. Hinges, pin holes and clamps all have assembly clearance, and clearance under repeated load only grows. The macroscopic symptom is that free play at the handles increases noticeably between the start and the end of the durability programme. A fair number of samples never fracture at all, yet fail because looseness exceeds the acceptance range.
So when you read the ISO 11199-2 item list, do not stop at the words "static strength" and "fatigue". Follow the load path and ask which region each item is interrogating. Loading points, loading directions, cycle counts and acceptance limits are as given in the current valid version of the standard text.
Brakes: where rollators lose points
Start by separating the two braking functions. One is the deceleration brake applied while walking, by pressing or lifting the handle lever, where the interest is whether deceleration is controllable, whether release is complete once the lever is let go and whether any residual drag remains. The other is the parking brake engaged by pushing the lever into a locked position, where the interest is whether the device moves on a slope once locked and whether it rolls away when the user sits on the seat. The test logic differs, and so does sample preparation.
A handful of failure modes recur across projects.
- Cable creep. The steel inner core stretches under tension while the plastic outer casing shortens under compression, so after repeated operation the overall brake travel grows and the parking lock position set at the factory drifts with it. The classic signature is a brake that passes before the durability programme and fails on re-test afterwards, with no visible damage to the brake itself. Listing brake cables as spare parts shipped with the samples saves half a scheduling cycle compared with sending them later.
- Mismatched friction pair. Solid rubber, PU and pneumatic tyres differ substantially in surface hardness and friction behaviour, and the same brake assembly on a different wheel can reverse the conclusion. If one model comes in several wheel configurations, state before submission which is the primary configuration and whether the rest count as variant models.
- Wet and contaminated conditions. Friction conditions degrade once the braking surface picks up water or dust. Where the standard calls for wet braking performance to be assessed, tread pattern, drainage path and friction material all become direct variables and dry-condition data cannot be extrapolated.
- Left-right asynchrony. When cable length or pre-tension differs side to side, one side bites first and the device pulls to that side during braking; on a slope it shows up as single-side slip that rotates the device. This one is easily waved through on the assembly line as "a difference in feel".
How height adjustment, folding and braking interact
The general engineering analysis of height adjustment mechanisms themselves, that is, pin hole wear patterns, clamp slippage and the spread of assembly tolerance across samples, is more usefully explained with the walking frame as the reference case and we cover it in the walking frame article rather than repeating it here. What a rollator needs on top of that is another layer: its handles are simultaneously load-bearing members and brake controls, so adjusting height and folding the device both rewrite the geometry of the brake system.
Height adjustment changes the free length of the cable. As the handle upright rises, the cable routing inside and outside the tube is drawn out and the free travel at the brake lever changes with it. Near the upper limit the cable may already be pre-tensioned, which shows up as residual drag after release; at a low setting the slack accumulates in the bends of the conduit, travel lengthens and the parking lock position moves forward. A brake set at mid height in the factory is not necessarily still within specification once the user changes height. So before submission, state which height the primary configuration uses, at which height the braking items are run and whether verification is needed at both ends of the adjustment range. Without those agreements the scope of validity of the report cannot be stated clearly.
The folding axis and the cable routing fight each other. A central X-fold makes the two uprights move relative to one another, and the cable has to cross the folding axis. Through repeated folding and unfolding, the cable at the bend experiences bending fatigue rather than tensile fatigue, and the failure location is usually at the end of the sheath rather than mid-core. If the durability items run the folding action and the braking action separately, that coupling is exactly what gets missed, and it is something the user performs every day.
The folding latch and the brake share a positioning datum. When the folding latch is not fully home, the relative angle between upright and base frame is off, and the contact position of the brake pad against the wheel shifts with it. In user language this becomes "every time I fold it I have to readjust the brake". This class of problem is invisible in single static tests and only reproduces in the fold-unfold-brake sequence, which is why a functional check should be repeated after the folding durability item.
On sample quantity, rollators have an extra reason to send more. The brake durability and parking items only mean something if they run in sequence on the same sample, while destructive strength items consume samples. Sending several units of the same batch and splitting destructive and non-destructive items across different samples both preserves the integrity of the braking chain and makes the conclusions more representative of the model.
Extra considerations for seated rollators
The seat does not change which part applies, but it does add a use scenario. When the user sits on the seat and leans back, the centre of gravity of the person-and-device system moves rearward. If the parking brake is not engaged, or its holding force is insufficient, the device rolls backwards; and if the combination of seat height and rear wheelbase already leans towards the unstable side, the tendency to tip rearward increases further. That is a qualitative conclusion drawn from static equilibrium rather than statistical data, but it makes one thing clear: the warning in a seated product's manual to engage the brakes before sitting has to be read together with the test conclusions, and document review will check for it.
Straps, storage baskets and cane hooks all need to be declared as well. A loaded basket shifts the centre of gravity of the whole device, and a hook takes an off-centre lateral load; both can affect stability and strength conclusions. An accessory that is present on the prototype but absent from the declared configuration list is a very common document inconsistency, and it usually surfaces only just before the report is issued.
What to prepare for submission
| Item to prepare | Detail | What happens without it |
|---|---|---|
| Samples | Several units from the same batch, primary and variant configurations clearly separated | With only one unit, destructive items leave nothing to re-test and the conclusion is thinly supported |
| Wear spares | Brake cables, friction pads, handle grips, castors | Damage mid-programme means waiting for a shipment and the whole schedule slips |
| User manual | Including maximum user mass, assembly and adjustment method, brake use warnings | Loading conditions for some items depend on manual declarations; a gap stalls the pre-test review |
| Structural drawings and material list | Tube specifications, weld locations, latch material and surface treatment | Failure analysis has nothing to work from and corrective advice stays at the level of describing symptoms |
| Configuration difference statement | Wheel types across the series, different height adjustment ranges, seated and non-seated | You may end up needing several reports, or being asked to add testing later |
| Height and folding state declaration | Primary test height, folding latch method, the state in which braking items are run | The scope of validity of the report cannot be written clearly and gets challenged during customer review |
| Target market information | Countries or regions you intend to export to | The item list may be scoped too narrowly and require later rework |
The item most often underrated here is the drawings and material list. If a crack appears during testing and the weld locations and base material grade cannot be matched to a drawing, failure analysis can only report "a crack at a certain location", which does not help redesign. Projects that hand the laboratory drawings in advance often get obvious weak points identified before formal testing begins.
Sequencing the tests so samples are not wasted
There is a general principle for sequencing: run the items that do not change the state of the sample first, such as overall dimensions, mass, functional checks and initial brake performance; then the durability and fatigue items; and leave the static strength items that load to destruction until the end. The reasoning is direct. A sample that has been through fatigue testing can still be used to verify whether the brake remains effective after durability and whether looseness in the latches has gone beyond the limit, whereas a sample loaded to destruction is only fit for scrap.
Another point easily overlooked is environmental conditioning. Metal, plastic and rubber parts respond to temperature and humidity at very different rates, and taking a sample straight from the shipping carton onto the test rig does not give the same result as conditioning it as the standard requires first. Plastic latch components and tyre-related items are particularly sensitive. Conditioning conditions and hold times are as given in the current valid version of the standard text.
If you also have crutch-type products to schedule, plan them together and set the sample batches against the scope in walker and crutch testing; previous scheduling approaches and failure analysis examples are set out in our testing case notes.
A few boundary questions
Does passing ISO 11199-2 mean I can sell? No. ISO 11199-2 provides safety and performance requirements and test methods. Market access depends on how the regulatory system of the target country or region is set up. A test report is part of the access material, not access itself.
Do I need a premarket notification for the US? Do not assume, and do not budget and schedule as if a submission is always required. The sound approach is to determine the product classification code that matches your specific construction, then look that code up in the FDA device classification database to see the class and whether an exemption exists. Class and exemption status follow the current database result; different construction or different claims can lead to different conclusions, so do not borrow someone else's answer.
Can several models in a series share one report? It depends on whether the differences affect the performance under test. Wheel type, frame tube specification, latch method, brake construction and height adjustment range are key differences and usually need separate verification; colour, grips and storage baskets can normally be covered as variant models. The more specific the difference statement, the more room there is to discuss.
A sample deformed during testing but did not break. Does it pass? That depends on what the standard says about residual deformation and retained function. Most strength items look not only for fracture but also at whether function remains normal after unloading and whether connections have come loose. This is another reason to run a functional check both before and after the strength items.
What we can do
SUNGO Mobility Testing Lab has long worked on wheelchairs, mobility scooters, walking aids and crutch-type products. We can take on testing under the ISO 11199 series, including ISO 11199-1, ISO 11199-2 and ISO 11199-3, and support failure analysis and verification after corrective action. Single-arm walking sticks are handled under their own applicable standard; which one applies and which edition to run is confirmed against the current valid version of the standard text. The laboratory is accredited by CNAS, CMA and IAS (USA), with laboratories in Shanghai and Hefei, so samples can be sent to whichever site is closer. Please note that accreditation marks demonstrate technical competence within the accredited scope and are not a commitment regarding market access outcomes.
A fuller scope is set out under testing services. If you already have a prototype and drawings and want to work out which items apply, how many samples to send and how long a report takes, call +86 132 4819 8029 or request a quote with your product documentation attached, and we will come back with an item list and a schedule proposal for your target market.