Draw the line between ISO 7176-5 and ISO 7176-7 first
"Just measure the dimensions of the wheelchair for us" is a sentence we hear constantly, and it forks the moment it reaches the laboratory. Overall vehicle dimensions run under ISO 7176-5, seating and wheel dimensions run under ISO 7176-7, and the domestic adoptions of both sit in the GB/T 18029 series. The two are not nested inside one another; they divide the work.
ISO 7176-5 answers the question "how much space does this chair take, what does it weigh, will it get through a doorway and into a lift". It measures the overall envelope in a defined configuration, the envelope when folded or dismantled, the space needed to turn and to reverse direction, and the mass of the complete chair along with the mass of removable parts. ISO 7176-7 answers a different question: "will a person fit on it, and where are the drive wheels mounted". It measures the geometric relationships among seat surface, backrest, armrests, legrests and footrests, plus the diameter, width and mounting position of drive wheels and castors.
The trap sits right here. A purchaser writes "supply a dimensional test report" into the technical agreement; the manufacturer measures the overall envelope, submits it, and gets it sent back, because what the purchaser actually wanted to cross-check was seat width, seat depth and backrest height, all ISO 7176-7 territory. The reverse happens too: a rehabilitation buyer tendering on accessibility receives a seating geometry report instead. Establishing "whose question is this report going to answer" before the order is placed saves more time than negotiating price.
The hard part of ISO 7176-7 is not the tape measure, it is the reference frame
Plenty of people assume seat dimensions are a matter of running a tape over the chair. Anyone who has actually done the work knows most of the effort goes into establishing the reference datum. Measurement under ISO 7176-7 is not read off the surface of the upholstery. Rigid plane boards as specified are first brought into contact with the seat and the backrest; those two boards define the seat reference plane and the backrest reference plane, their intersection gives the reference line, and it is inside that coordinate system that seat depth, seat width, effective seat surface angle, backrest angle, armrest height, footrest-to-seat distance and the rest are defined.
Get the datum wrong and every number downstream shifts together. Three failure modes recur.
The first is not waiting for the upholstery to settle. Fabric covers and high-resilience foam keep deforming slowly after the board is applied, so an operator who places the board and reads immediately gets a different answer from one who waits for deformation to stabilise. The disciplined approach is to agree a stability criterion before reading; the applicable loading conditions and hold requirements are as given in the current valid version of the standard text.
The second is board placement that varies with the operator. Neither seat nor backrest is a true plane. Most products have a seat that dips from front to rear and a backrest with a lumbar prominence. Where the board is placed and which section it contacts varies with operator habit, and reproducibility scatters accordingly. A mature laboratory writes the placement rule into its internal work instruction and keeps photographic records.
The third is conflating "seat dimensions" with "user dimensions". ISO 7176-7 measures the geometry of the product, not anthropometric data. The body sizes quoted in a user manual are the manufacturer's declaration. The two sets of numbers mean different things and cannot substitute for each other in a report.
Which dimensions are taken loaded and which unloaded
The previous section settled where measurement starts from. There is an equally important question that gets taken for granted: is a given item measured on an empty chair, or with load applied first?
Two opposite misconceptions are common. One holds that since a seat is made to be sat on, everything must be measured loaded. The other holds that a dimension is a dimension, so loading is irrelevant. Either reading will leave the manufacturer's in-house numbers disagreeing with the laboratory report. ISO 7176-7 does not issue a blanket "all loaded" or "all unloaded" rule; it specifies the measurement state item by item, with some quantities taken unloaded in a defined configuration and others taken under load.
To understand where the line falls, ask whether the quantity in question would be changed by occupant load.
Quantities that tend to be taken unloaded are geometric relationships between rigid structures. Wheel diameter and width, track and wheelbase, castor fork hole positions and offset, armrest tube lengths, the relative positions of mounting points on the frame: these are set by metal and hard components, and whether or not something presses on them makes little difference. Applying load does not bring the numbers closer to real use; it introduces new sources of uncertainty such as tyre contact flattening and bearing clearance being taken up.
Quantities that tend to be taken loaded are those where a soft interface participates. Effective seat depth, seat surface angle, backrest angle and the real distance from seat to footrest are all affected by cushion foam compression, backrest fabric tension and suspension travel. Unloaded, you read "the shape of the product at rest"; loaded, you read "the shape with someone sitting on it". Both numbers are true, but they answer different questions.
It has to be said plainly that the above is a way of understanding the boundary, not a measurement list you can copy. Which state applies to each specific item, how load is applied, what counts as stable and when the reading is taken are all as specified in the current valid version of the standard text; for domestic submission, cross-check against the corresponding adopted documents in the GB/T 18029 series. Doing that comparison before in-house measurement is far cheaper than arguing afterwards about who measured correctly.
What genuinely belongs in your internal rules is this: every dimension in the report must state the loading state under which it was obtained. If unloaded, say unloaded. If loaded, state how and where the load was applied. Only then can a purchaser holding two reports from two laboratories tell whether the difference comes from the product or from the measurement state, and only then does a manufacturer have a stable baseline for re-measurement after a design change. Most cases of "the numbers do not agree" turn out, when traced, to be two parties measuring the same named quantity in different states.
The loading apparatus is the test dummy specified in ISO 7176-11 (this article series uses the term "test dummy" throughout; the corresponding domestic documents sit in the GB/T 18029 series). Test dummies come in different size bands, and which band to use, how it is positioned and how it maps onto the declared user range is a separate topic that we cover in a dedicated article. One point is enough here: pick the wrong band and the entire loaded data set is void, and the cost of redoing it far exceeds the cost of one extra question before submission.
There is also a sequencing detail worth agreeing up front. Running loaded measurement before unloaded measurement on the same prototype, versus the other way round, can shift the unloaded numbers, because foam and fabric retain residual deformation. Fix the measurement sequence at the planning stage and record it in the report rather than leaving it to whoever is on the bench.
Adjustable mechanisms: without an agreed configuration the report cannot be used
Modern wheelchairs keep adding degrees of adjustment: backrest angle, seat angle, stepped seat height, stepped armrest height, continuously adjustable legrest angle, rear axle position moving fore and aft as well as up and down, castor forks with alternative hole positions. ISO 7176-7 data are taken in one specific configuration, and a different configuration gives different data.
So two things must be answered before submission. First, what is the nominal configuration, meaning the factory default state. Second, do the endpoints of the adjustment range need to be covered? Measuring the nominal configuration alone gives a compact report, but it leaves you unable to answer the purchaser who asks what seat depth remains with the backrest reclined to its limit. Covering the endpoints raises effort and cost, but produces a report that supports the specification sheet directly.
The awkward case in practice is a manual that never states the adjustment range at all. The manufacturer feels that "it adjusts, obviously", the laboratory has no basis to work from and defaults to a mid position, and neither side is satisfied with the result. A declared adjustment range is one of the lowest-cost documents in the submission pack and one of the highest returning; a single page can save a full round of rework.
Wheel dimensions: the section that gets treated as trivial
Wheels look simple. Measure the diameter, measure the width, measure track and wheelbase. In practice it generates plenty of disagreement.
Tyre condition is the main source. The outside diameter of a pneumatic tyre varies with inflation state, so pressure must be set to the state marked by the manufacturer before measuring, and the source of that basis recorded in the raw data. Where the marking is unclear, the laboratory will ask the manufacturer to confirm in writing; the applicable inflation conditions and tolerances are as given in the current valid version of the standard text. Solid tyres avoid this issue, but where the same model is supplied with both pneumatic and solid tyres, the two configurations cannot share one set of geometric data. That is one of the more common reasons a dimensional report is rejected.
Castor mounting geometry also has to be stated. Castor fork offset and the relationship between castor axle and steering axis directly influence self-centring behaviour and rolling resistance. Some products have multiple mounting holes in the fork, and changing the hole effectively changes the handling character of the chair, so the hole position must be locked in the measurement record.
The same applies to drive wheel axle position. Moving the rear axle forward makes the chair more manoeuvrable and less stable; moving it back does the opposite. ISO 7176-7 records the geometric position, while stability is dealt with in another part of the GB/T 18029 series, but the prototype configuration behind both reports must match, otherwise the pair contradicts each other. A reviewer who sees a forward axle in the dimensional report and a rearward axle in the stability report will usually ask for both to be redone.
One table for scope and preparation
| Question you need answered | Primary basis | Typical measurements | Settle this before submission |
|---|---|---|---|
| Will the chair fit a lift, a doorway, a corridor turn | ISO 7176-5 / GB/T 18029 series | Overall envelope, folded envelope, turning space, mass of chair and removable parts | Folding method, whether parts are removed, which parts |
| Does the seating fit, where do specification sheet seat width and depth come from | ISO 7176-7 / GB/T 18029 series | Seat depth, seat width, backrest height and angle, armrest height, footrest to seat distance, seat angle | Nominal configuration, whether adjustment endpoints are measured, cushion configuration |
| Is a given item measured loaded or unloaded | ISO 7176-7, loading apparatus per ISO 7176-11 | Item-by-item measurement state, method and position of loading | Which items require load, how the loading state is recorded |
| How are drive wheels and castors recorded | ISO 7176-7 | Wheel diameter and width, track, wheelbase, castor offset and hole position | Tyre type, basis for inflation state, list of optional hole positions |
| Will the ISO 7176 reports contradict each other | ISO 7176 series as a whole | Consistency of prototype configuration across sub-reports | A single written prototype configuration declaration |
Where data scatter comes from, seen through the load path
What follows is a qualitative deduction from structural loading and material behaviour, not a statistical result from any particular batch of samples. Its only purpose is to help you decide where to add control.
Scatter in seat depth and backrest angle comes mainly from the time-dependent response of soft interfaces. Foam under compression is viscoelastic and its stress decays with hold time, while fabric is a flexible tensile member whose tension distribution depends on how tightly it is anchored. Together they mean that "how far to press before it counts as stable" is itself a criterion someone has to define. Different definitions, different numbers. Writing the stability criterion into the work instruction is more effective than debating afterwards who measured more accurately.
Scatter in armrest height and footrest distance comes more from accumulated assembly tolerance. These parts connect to the main frame through tubes, clamps and pins, stage by stage, and each fit clearance is taken up to one side under self-weight and load. Remove and refit the armrest on the same chair repeatedly and the reading moves within a band. The engineering answer is a defined refitting procedure, not a demand that the operator measure more carefully.
Scatter in wheel geometry is governed largely by tyre contact deformation and bearing clearance. Under load the contact patch flattens and the hub-to-ground distance shrinks, so taking the reference point at the rim rather than at the hub centre leads to conclusions that point in different directions. That is why wheel data must always be reported together with the definition of the reference point; the number alone means nothing. This also explains, from another angle, the earlier point about loading state: a dimension without a stated loading state cannot be reconstructed by the reader.
What to do once you have the report
For domestic sales, the data corresponding to ISO 7176-7 usually go into the product specification and accompanying documents for the purchaser to check, with the matching domestic documents in the GB/T 18029 series. For export to the EU, seat and wheel geometry data are typically part of the supporting material describing product characteristics and intended users in the technical documentation, filed alongside the other ISO 7176 sub-reports; what has to be submitted and which text applies is governed by the currently valid regulations and standard texts.
The US route needs to be assessed separately. For this kind of product, look up the specific product code in the FDA classification database to confirm the class and whether 510(k) is exempt. Do not assume across the board that a submission is required, and do not assume across the board that it is exempt; the database result governs. Keep a record of the query outcome as the basis for later discussion.
One more reminder: the validity of a dimensional report is tightly bound to the prototype. Changing the cushion supplier, changing the backrest fabric, redesigning castor fork hole positions, any of these can leave the original report inconsistent with what is in production. Asking "will this change touch any quantity in ISO 7176-7" at the design change stage is far cheaper than having a reviewer find it after mass production has started.
How we can help
SUNGO Mobility Testing Lab is the dedicated wheelchair and mobility aid testing laboratory within our group, working regularly on wheelchairs, powered wheelchairs, mobility scooters, walkers and crutches. We can carry out seat and wheel dimension measurement together with overall dimension and mass measurement to the ISO 7176 and GB/T 18029 series, and support you on the prototype configuration declaration, adjustment range confirmation, loading state agreement and measurement plan design, so that the several reports produced on one prototype stay consistent with each other. The laboratory is accredited by CNAS, CMA and IAS (USA), with laboratories in Shanghai and Hefei. Please note that accreditation marks demonstrate technical competence within the accredited scope and are not a commitment regarding market access outcomes.
If you want to know how the items would be scheduled, how many prototypes to prepare and which documents to have ready, start with our testing services and standards index, or look directly at the scope notes for manual wheelchair testing and powered wheelchair testing. Call +86 132 4819 8029 or request a quote and send us the product documentation; we will start by telling you which items you actually need.