doubt recertification c

[GHIZMO]

I would have three doubts regarding the addition of components to machines already certified

doubt 01
I wanted to ask if, in case of addition of mobile component within a completely segregated automated island, a recertification of the whole machine is necessary.
to better understand, the machine is certified by the manufacturer, if the user independently adds a component that performs a "simple" function like a cylinder that controls the movement of a sensor within the segregations without affecting the protections, this cause needs to remake all the actions of the case and then be able to draw up a certificate there?

_
Considering the same machine as before, if the user decides to add a component type a unplanned exhaust robot in the original machine there are differences if the discharge is made using an exhaust path already present type a slide or a palletizer that involves openings in the protections.

doubt 03
If an external component is added to the machine not connected to the cycle level to it type an exhaust tape or a palletizer that I can operate regardless of whether the machine is active or not is considered part of the machine and therefore I must re-certificate?

at the end of all, how do you deal with the technical file in case of recertification since the user naturally does not have the one made by the manufacturer (and even can you ask)?

Thank you.

 

[gerod]

01.
depends, but I would say if you change the use destination or change the mode of operation I would say yes.
It is obvious that the manufacturer of the island, in case of an accident, will tell you: you modified the machine, and then it is your business (always that on the manual is written! )

02.
always implies a c iter, at least of complex machine (constituted by robot + existing machine).
The robot is not machine, it is almost machine and therefore does not comply with all the res of the directive.
then, opening or not opening the protections you must certify the two units.

03.
No, if the machine is outside it doesn't need.
It is obvious that then the machine + tape could always be a complex machine, so iter ce.

on the last question: you do what you can, you make an integration and a new analysis of the risks.

a greeting

 

[Fulvio Romano]

02.
always implies a c iter, at least of complex machine (constituted by robot + existing machine).
The robot is not machine, it is almost machine and therefore does not comply with all the res of the directive.
then, opening or not opening the protections you must certify the two units.

and if the risk analysis does not show any risks other than the original machine?

 

[meccanicamg]

for the fact that it changes what provided by the plant supplier it is definitely necessary to recertificate us when what is added has not been evaluated by the supplier and the responsibility falls on who has modified the plant, with charge of responsibility.

It is also true that if the final clinete wants to add or change internally or externally the plant can do self-certification thus combining the original system with the declaration of incorporation or (depending on whether machine or quasimachine) of the added part. is more practical, fast and costs less since self-certification is to be charged to itself, that is to the user customer himself.

 

[gerod]

and if the risk analysis does not show any risks other than the original machine?

attention: residual risk is not synonymous with non-compliance.
here the question is: I have a car a and I make it work with a car (almost car) b.
the a+b set, while being marked there and b not marked ce (the almost machines are not in conformity with the machine directive), to be there, needs a iter ce, risk analysis, etc.
minimally there is a danger that the initial manufacturer of the machine had not foreseen: the interaction between machines.
In short, if one of the machines is not machine (almost machine) you have to make the iter ce.

the same speech is worth if the two machines are there. a c + does not make a c final, as a rule.
I hope I have clarified

It is also true that if the final clinete wants to add or change internally or externally the plant can do self-certification thus combining the original system with the declaration of incorporation or (depending on whether machine or quasimachine) of the added part. is more practical, fast and costs less since self-certification is to be charged to itself, that is to the user customer himself.

I'd say it's reductive. not only self-certification but also the drafting of the technical dossier of the construction that goes to integrate machine and almost machine.
the ddc is done after checking the res.

 

[meccanicamg]

the ddc is done after checking the res.

I think so, theoretically, and who knows what the word security means. the majority of companies do not make risk assessment, they do not have complete documentation neither for single components nor for large standard series (manual of use, electric schema and/or tire si e nò...not always, risk assessment almost never, technical dossier incomplete, errors in the drafting of the ddc with references to wrong norms or no more in force). I've seen them cooked and raw and all together.

 

[gerod]

I've seen them cooked and raw and all together.

more than I would say "not seen" since only 5% seems to do things well.
of cooked and raw?
The last one I saw: a paste copy manual that little described that machine.
ocio, and in other:
.... suspension points on a chapter dedicated to safety as to create some attention!
the beautiful is that this manual is relative to an island and the manufacturer of the manco island has seen it (he has integrated his manual but obviously has read little)

 

[pollo]

...I've seen them cooked and raw and all together.

I the most beautiful I saw was:
the machine complies with all past, present and future directives.
blessed them, I always miss some res