Subscribe to our newsletter to receive the latest news and events from TWI:

Subscribe >
Skip to content

CTOD Testing


The concept of fracture toughness was introduced in an earlier Connect article, Job knowledge 71, which discussed the Charpy-V test, a simple qualitative test that gives only an indication of the toughness of a metal.

The next few articles will look at the tests that enable fracture toughness to be accurately measured in a quantitative manner by using a full size specimen containing a crack with loading that is representative of service conditions.

This allows a fitness-for-purpose analysis to be carried out which enables a critical defect size to be calculated. Thus, prior to fabrication, realistic acceptance standards can be set and decisions on appropriate NDE techniques and detection sensitivities can be made.

For equipment already in service, it is possible to justify the continued use of cracked or otherwise flawed components until such time as repair or replacement can be effected. Such engineering critical assessments can save an operator large amounts of time and money, running into perhaps hundreds of millions of pounds in the case of an oil rig for example. Whilst the Crack Tip Opening Displacement (CTOD) test was developed for the characterisation of metals it has also been used to determine the toughness of non-metallics such as weldable plastics.

The CTOD test is one such fracture toughness test that is used when some plastic deformation can occur prior to failure - this allows the tip of a crack to stretch and open, hence 'tip opening displacement'.

Unlike the inexpensive 10mm by 10mm square Charpy-V test piece with a blunt machined notch, the CTOD specimen may be the full thickness of the material, will contain a genuine crack and will be loaded at a rate more representative of service conditions. Conventionally three tests are carried out at the relevant temperature to ensure consistency of results.

The test piece itself is 'proportional' - the length, depth and thickness of each specimen are inter-related so that, irrespective of material thickness, each specimen has the same proportions.

There are two basic forms - a square or a rectangular cross section specimen. If the specimen thickness is defined as 'B', the depth (W) will be either B or 2B with a standard length of 4.6W. A notch is machined at the centre and then extended by generating a fatigue crack so that the total 'defect' length is half the depth of the test piece- see Fig.1. A test on a 100mm thick weld will therefore require a specimen measuring 100mm thick, 200mm wide and 920mm long - an expensive operation, the validity of which can only be determined once the test has been completed.

Fig.1. Proportional rectangular cross section CTOD specimen
Fig.1. Proportional rectangular cross section CTOD specimen

The test is performed by placing the specimen into three point bending and measuring the amount of crack opening. This is done by means of a strain gauge attached to a clip placed between two accurately positioned knife edges at the mouth of the machined notch (Fig.2)

Fig.2. Typical test arrangement. The specimen can be easily immersed in a cooling bath
Fig.2. Typical test arrangement. The specimen can be easily immersed in a cooling bath

As bending proceeds, the crack tip plastically deforms until a critical point is reached when the crack has opened sufficiently to initiate a cleavage crack. This may lead to either partial or complete failure of the specimen. The test may be performed at some minimum temperature eg the minimum design temperature or, more rarely, at a range of temperatures.

The values that are required for the calculation of toughness are firstly the load at which fracture occurs and secondly the amount by which the crack has opened at the point of crack propagation (Fig.3).

Fig.3. Position of CTOD specimen immediately prior to crack propagation
Fig.3. Position of CTOD specimen immediately prior to crack propagation

Since the length of the crack and the opening at the mouth of the notch are known it is a simple matter to calculate the crack tip opening by simple geometry. Whilst the test is in progress the results are recorded automatically on a load/displacement chart (Fig. 4). This illustrates the various shapes of curve that may be produced - (a) is a test where the test piece has fractured in a brittle manner with little or no plastic deformation. (b) exhibits a 'pop-in' where the brittle crack initiates but only propagates a short distance before it is arrested in tougher material - this may occur several times giving the curve a saw tooth appearance or after this one pop-in deformation may continue in a ductile manner as in (c) which shows completely plastic behaviour.

Fig.4. Load vs crack opening displacement curves showing three types of fracture behaviour
Fig.4. Load vs crack opening displacement curves showing three types of fracture behaviour

The location of the notch in the weld HAZ or parent metal is important as an incorrectly positioned fatigue crack will not sample the required area, making the test invalid. To be certain that the crack tip is in the correct region, polishing and etching followed by a metallurgical examination are often carried out prior to machining the notch and fatigue cracking. This enables the notch to be positioned very accurately. Examination may also be carried out after testing as further confirmation of the validity of the test results.

Once the sample is broken open the crack surface is examined to ensure that the fatigue crack has a reasonably straight front. The residual stresses present in a welded joint may cause the fatigue crack front to be irregular - if this is excessive the test may be invalid. To overcome this problem the test piece may be locally compressed at the machined notch tip to redistribute the residual stress.

Two depressions each side of the sample can often be seen where this compression has been carried out. The fatigue cracking itself should be carried out using a low stress range. The use of high stresses to speed up the fatigue cracking process can result in a large plastically deformed area ahead of the fatigue crack and this will invalidate the results of the test.


Other causes of test failure can unfortunately only be determined once the test has been completed and the crack surface examined. The precise length of the fatigue crack is measured - this is required for the analysis - but if the length of the crack is not within the limits required by the specification the test is invalid. If the fatigue crack is not in a single plane, if the crack is at an angle to the machined notch or if the crack is not in the correct region the test may need to be repeated.

Related specifications

BS 7448 Parts 1- 4 Fracture Mechanics Toughness Tests
BS 7910 Guide on Methods for Assessing the Acceptability of Flaws in Metallic Structures.

ASTM E1820

Standard Test Method for Measurement of Fracture Toughness. 


15653 Metallic Materials - Method of test for the determination of quasistatic fracture toughness of welds

This article was written by Gene Mathers.

For more information please email: