We have re-written sections 5。3 – 5。5 in a more succinct manner。

Please find below a point by point response to the comments of the reviewers。

Point by point response to reviewers’ comments。

We like to thank the reviewers for the very valuable comments。 We have studied them very carefully and we have carried out the necessary revisions。

Reviewer A

Main Points:

The focus of the paper is discussed on page 2 in section 1, stating the need to test scaled models and the use of Buckingham’s Pi theorem to develop the similitude relationships between the prototype and the model。 More information on this, with the similitude relationships for the ½ scale ROPS model treated in this paper, can be found on page 3 in section 2。

We are sorry for not giving some of the information at the very start of the paper。 Details of the ROPS model, using hollow sections, are now given in section 2。1 on page 3。

We thank the reviewer for pointing out a typing mistake on the cross section dimensions of the model in Table 2。 We are sorry about this。 This has now been corrected, with proper values of the width and depth of the cross-section。 The second moments of area of the model

and proto type satisfy the required similitude relationship of IM = IP/16, where M and P denote the model and prototype respectively。 This is discussed in section 2。1 on page 3。

Additional Points:

The roof canopy shown in Figure 1 is to provide protection under falling objects。 This is a detachable structure, which is omitted in the present treatment。 Only the ROPS frame shown in Figure 2。2 is considered。 This is explained in section 2 on page 2。 In Figure 3, two opposite views of the ROPS are shown in order to illustrate all the strain rosettes and LVDTs。 This is explained section 3。1 on page 5。

The scaling factor of ½ is used for all the cross-section dimensions。 As mentioned above, there was a typing error in Table 2 and we are sorry about this。 This has been corrected。 This is discussed in section 2。1 on page 3。

During experimental testing, it was observed that there was some change of cross- sectional shape at the regions of plastic hinge formation。 This was also observed in the FE analyses of the ROPS models。 This is explained in sections 3。2, 4。6。1 and 5。5。

Von Mises stress is derived from the strains using standard equations found in text books。 We have made reference to this in section 3。2 and quoted the book by SJ Moy。

The vertical loading phase took place after the lateral loading phase, with residual plastic strains and deformation in the ROPS。 This is evident from the approximately 5 mm initial deflection (starting point) of the loading curves。 The experimental and FE curves do not compare well due to the sustained plastic deformation that took place after the initial lateral loading phase and the inability of the FE model to replicate this properly。 This is explained in sections 4。6。2 on page 11。 We have shown the loading and unloading curves in Figure 6。

We have clarified what 10mm and 5 mm mesh densities mean。 These are the dimensions of the finite elements in the 2 models, as explained in section 4。1 on page 8。

Section 4。6:

1。(new) page number 10, the sentence “----underwent significant deformation”, is not clear。 We have improved this。

2。first paragraph of section 4。6。2 , “It is evident that –the reduced stiffness is inherent---“ – is not clear。 This is because the ROPS has already suffered plastic deformation in the previous lateral loading phase, as explained in the same section。

3。section 4。6。3, on new page number 12 ,“-----did not compare well”。 Why? The main reason was because the FE model was unable to replicate the sustained plastic deformation in the ROPS that took place during the initial lateral loading phase。 We have explained this in the sections 4。6。2 and made reference again in section  4。6。3  on (new) page number 12。