4。last paragraph of section 5。5 is not clear。 We have clarified this in section 5。5, on new page number 17。 Two of the load deflection plots (for the stiffer ROPS) have higher loads and smaller deflections, while the other two plots have the opposite trends。

Sections 5。3 – 5。5 have been re-written in a more succinct manner。

Conclusion: we have addressed the suggestion provided by the reviewer and we have discussed the background, aims and procedure of the paper in the last paragraph in section 1 on page 2。 The flow of the paper is compatible with this

Reviewer C

We have described the background of our paper, its aims and procedure in the last paragraph of section 1 on page 2。

Suggestions for Providing Additional Details:

1。A summary of the Australian Standard AS2294-1997 has been included in the second paragraph of section 1 on page 2。

2。Description of the test set up and more details of testing are provided, together with an additional Figure (2。1) of the ROPS testing frame in sections 3 and 3。1 in pages 3, 4 and 5。

3。Justification of similitude modelling, especially for load – deflection response under lateral load is discussed in section 4。7。 Figure 11 confirms the similitude relationship between model and prototype for lateral deflection (and energy absorbed through the area under the curves)。

4。We are not absolutely certain of the slight kink in the stress vs deflection response in Figure 5。 It could be due to experimental error in strain recording which has resulted in the kink when converted to stress。 But, this slight kink has not negated the usefulness of this Figure which clearly shows that there is yielding at the location。

5。We do not have the predicted strains。 But the predicted stresses from the FE model agree very well with the stresses obtained from the measured strains at the locations on yielding (sections 3。2 and 4。6。1)。 Moreover, the predicted and measured lateral load vs deflection responses (Figure 4) agree very well, giving further confidence in the computer (FE) modelling。

6。We have replaced the word” calibrated” by the word validated and explained that this was for the lateral load response of the ROPS, which is further investigated in the paper。 This is discussed in section 4, paragraph 2。

7。We have given details of the FE model, including the stress-strain relation, which can be used by others to replicate the results。 This is given in sections 4, 4。1, 4。2, 4。3 and 4。4 in pages  8, 9 and 10。

8。We have improved the legend in Figure 9 as requested。

Technical contribution of our work:

In addition to understanding ROPS behaviour, we have established the feasibility of using FE techniques for ROPS evaluation and design

1。We validated the FE model of the ROPS for investigation

As mentioned in section 1 of the paper, the main aim of this paper was to establish the feasibility of using FE techniques for design and evaluation of ROPS。 Towards this end, we have successfully validated the FE ROPS model for studying lateral load response, which is the most common type of roll over and which was the focus of this paper。 The other two loading sequences are strength cases to ensure that a ROPS has adequate capacity in the vertical and longitudinal directions。The lateral load deflection profile from FE analysis compared very well with that from experimental testing。 Both the experiment and the FE analysis showed yielding at the top and bottom of the posts under lateral loading of ROPS。 Moreover, similitude verification provided confidence for using FE models of full scale ROPS in further investigation。 We feel that we have provided sufficient details for validating the FE model for studying lateral load response。

2。FE techniques can be used in ROPS evaluation and design。