Direction of test for fire doors

(These notes are compiled by DJ Streicher (Pr.Eng.) of Ignis Testing Cape Town November 2020)

From the fire through the approval of a building to the design and testing of materials, decisions have to be made. It starts from identifying materials for purpose including the strength of materials and inter alia the properties of materials and specified building elements which will prevent and/or contain fire spread in buildings. Two design considerations are taken into account for fire spread design namely to protect people and/or property. This will include early warning systems, active fire suppression systems, and passive resistance of building elements of which fire-doors contribute a large and critical proportion.

The problem starts to identify from which direction a fire will come. If an element is completely symmetrical, it is common sense that the direction of the material relative to the fire direction does not matter. If the element is non-symmetrical then it does matter to know whether the element will have the same resistance against fire from either direction. It is thus clear that in this event a non-symmetrical element must be tested bi-directionally to determine which direction is most onerous. If fire resistance tests of elements show no difference with respect to direction, then they can be designed the same way as asymmetrical elements. If it differs, then fire designs must take this phenomenon into account so that the highest resistance is in the direction of the fire. Single-sided hinged fire doors react differently when tested bi-directionally. Why is this important? Because most divisional fire-door elements will consist of single-sided hinged doors.


Testing codes for doors consist usually of two primary codes. The first code guide the testing authority about equipment requirements, testing procedures, observations, and reporting. The second code will be a specific code against which parameters a door must be tested. SANS 10177-1 is a general administration code and SANS 10177-2 is the operational test code directing the testing authority on how to test. SANS 1253 is the fire doors and fire shutter code against which a door assembly will be tested. The equivalent European codes are EN 1363 parts 1 & 2 and the door code is EN 1634 parts 1 & 3. EN 1634-3 is for smoke control. The European codes are much more descriptive and address more parameters and test methodologies. There is no SABS smoke test control code for fire doors and shutters. When comparing codes it is evident that the SABS and EN codes differ substantially. For instance, the EN code limits the frame temperature which is not the case with the SABS code.
The time-temperature curve is a key parameter that is, fortunately, the same for all the fire resistance codes around the globe. It is this test input parameter that enables the comparison of doors from different manufacturers and countries. This so-called standard fire curve is in use for more than a hundred years and is the baseline reference for door test results in terms of resistance time. The outcome might be stricter for European results because additional items such as cotton pad tests, frame temperature requirements, radiation limits in kW/m², etc. could lead to higher standards. It must be noted that the insulation parameters are the same for the EN and SANS codes.


Clause 3.7 of SANS 1253 defines a fire-door assembly as a door and its essential hardware so mounted on a frame that the required fire resistance is provided. Clause 3.6 describes a fire-door. At this point, it must be noted that a door is an assembly inclusive of the frame and hardware. A door leaf is a part that hangs on the frame or is alternatively described as the moving part. There is thus no doubt that a fire door test is inclusive of the entire combination of frame, fixing, frame type, door leaf and essential hardware to be built into the test furnace to resemble the final installation of the fire door.

Clause 4.2.1 clearly states A fire-door assembly shall be of fire resistance A, B, C, D, E or F (see table 1), as required (see annex A). Clause 4.2.2 gives a list of types. We are only referring for this document to single- or double-leaf, hinged, single-action fire-door assemblies.
It is here where we have to define symmetry. Because the code refers to a fire-door assembly, symmetry applies to the assembly and NOT only to the leaf. A single action hinged door fitted in a rebate is non-symmetrical for obvious reasons such as the hinge pins are located on one side of the leaf, the rebate protects part of the leaf on one side and the door is not positioned in the centre of the frame.

Clause states that if the construction (face to face) is non-symmetrical, test TWO samples to establish the fire resistance from both sides and take the lower of the two results as the actual fire resistance. This clause overrides the direction clause in SANS 10177-2.


Clause 3.5 of SANS 10177-2 defines Exposure in the furnace. Clause 3.5.1 addresses positioning.
3.5.1 b) in the case of a test specimen of an element that has the function of separating spaces, heat will be applied over the whole of one face only, the selection of that face being as follows:

1) If the element will be required to resist fire in one direction only, the heat will be applied in that direction;
2) if the element may be required to resist fire from either direction, the heat will be applied in the direction considered, by the testing authority, to give the lower resistance (in cases of doubt, each face should be tested on separate test specimens);

There is no doubt that clause 3.5.1 b) of SANS 10177-2 acknowledges the fact that there could be a difference in the fire resistance of an element depending on the direction of testing. Tests according to SANS 1253 are tested according to SANS 10177-2 which directs and controls the testing procedures and methodology.

Also, note that the onus is on the testing authority to direct the sponsor which direction gives the lower resistance. At this point, it must be noted that if the sponsor instructs the testing authority to test a door in a direction giving lower resistance, then the testing authority will have to inform the sponsor as such. If a door is tested in the lower resistance direction, then according to 3.5.1 b) 1) it is accepted it is the direction in which the door will be installed. It means by implication, and in the spirit of the code, that the door cannot be classified as a door with the same resistance in the other direction unless the door is tested in both directions.


In terms of door testing the direction of testing defines which direction the door opens relative to the furnace exposure. It is most commonly referred to as open inwards or opens outwards with reference to the exposed side. What is important to note is that in the case of an open inwards door, the rebate is also on the exposed side.

Historic testing and test experience showed that there is a difference between the fire resistance of a door opening inwards to the fire, or a door opening outwards away from the fire. It is a known fact that an open inwards door is the more onerous direction of resistance. In short, it means the door assembly will fail quicker compared to an open outwards door. This phenomenon is well documented and accepted by testing authorities and international codes.

The EN 1634-1 code for fire doors has a specific clause, 13.4 Asymmetrical assemblies. This code acknowledges the principle for a difference between the direction of testing for doors. Table 2-Type of a door set and direction to be tested to cover the opposite direction covers the asymmetrical aspect. Only hinged doors with timber leaves opening inwards are covered in this table to include doors opening outwards w.r.t. resistance but, not the other way round. A metal leaf in a metal frame can be regarded as the same open outwards or inwards excluding insulation properties. In conclusion to this table, it is only a timber leaf in a timber frame that can be classified as an open outwards door if tested open inwards. All the rest have insulation limitations. Note that no door with a timber leaf opening outwards will be regarded as passing a test opening inwards.

Annexure C (Informative) of the EN 1634-1 code explains the Background to the field of direct application statements for asymmetric constructions and supporting constructions.


Rebate size is a moot point at many final building inspections. To clarify this point clause in SANS 1253 must be quoted. The width of the rebates of frames shall be at least 25 mm and the material thickness of steel frames should be at least 1.5 mm for class A, B, C, and D Doors. The first note to make is it does not apply to class E doors the way the clause is written.

Other codes do not specify dimensional parameters is because a fire-door code is a performance specification, not deemed to satisfy the specification. The question to ask is if a door assembly passes a test with a smaller rebate size, does this negate the test result? There is not enough evidence in all cases to prove that a larger rebate achieves higher fire resistance. There could even be cases where larger rebates are counterproductive.


A phenomenon observed was the closing of doors with barrel bolts for testing purposes instead of fitting the locksets or hardware which will be fitted on the production doors. We had test failures of well manufactured and designed doors because of incorrect lockset cut-outs. This type of failure would not be noticed if doors were fitted with bolts instead of the proper hardware. These tests failed before 20 minutes because of integrity failure at the lock. This emphasises the importance of testing a door assembly and not with the emphasis only on the door leaf.


There is a clear misunderstanding and misinterpretation of the direction of fire door testing results in South Africa. The SANS fire testing codes are very clear that testing results must be comprehensive and complete. SANS 10177-2 clause 4.2.4 states that additional observations must also be made and recorded.

If one analyses SANS 10177-2 and SANS 1253, then the conclusion is that according to SANS 1253 clause, two samples of single-action hinged doors should be tested in two directions to determine the lower resistance. The code further states clearly that the lower resistance must be
given for a fire-door if tested according to SANS 1253. This by implication means any classified fire- door can be fitted in any direction.

It must be noted that information obtained from prospective door suppliers revealed that doors tested by them in the past, were tested open outwards. This implies that most, if not all, tested doors in the past were tested open outwards. In the case of single-action hinged doors, this will not give the lowest fire resistance as required by the SANS codes. This way of testing was confirmed by an employee of the SABS fire testing facility.


It appears that a president is established to test doors open outwards and that the results are accepted by the fire chiefs and fire engineers as the fire resistance class irrespective of direction. It is required that the authorities, fire chiefs and fire engineers make a stand and declare if in future fire-door resistance results of open outward doors, will or will not be accepted as the fire resistance class irrespective of the direction of application.
There is no prior direction information for fire doors sold at general hardware outlets. This means that the lowest resistance must be applied to those doors because a general door can, and will be, applied in any direction. It appears that in South Africa, the direction of testing and reporting was not explicitly addressed and stated in the past.


There is a trend in the rest of the world to promote bi-directional testing of fire doors. This trend was sparked by the catastrophic failure of the Grenfell apartment block in London. The SANS fire testing codes nonetheless support the bi-directional testing and difference of non-symmetrical door assemblies. It does, however, seem that it is not adhered to by the testing authorities and is also not acknowledged by the approval authorities in South Africa. We are leaving the door open for legal action against the industry if the codes are not followed unless we waive the code requirements for bi-directional testing to test only from one side, whether it is open inwards or outwards.

The fire chiefs and authorities will have to address this point urgently and make the call.

Just Glass CC t/a Ignis Testing Reg no 1996/047240/23


‭(021) 919-4299‬ 

(021) 905 0322


20 Chardonnay Rd, Saxenburg, Blackheath, Kuilsrivier, 7580

Office Hours

Monday – Friday:

8am – 5pm