Hopp til innhold

Forsiden / Blogg / ANSYS-bloggen / New NORSOK Requirements for Drop Tests in 2013  

+47 67 57 21 00

+46 21 470 35 50

Kontaktskjema

New NORSOK Requirements for Drop Tests in 2013

From January 1st 2013 NORSOK R-002 is the governing standard for lifting equipment. It states that all portable offshore units shall be designed and tested according to DNV2.7-3. This means that either physical or simulated drop test must be performed if the PO unit is class R60&R45 or R60-SE. The reason for this new requirement is stated in DNV 2.7-3; “Impact loads may occur during lift off or set down of PO Units and they are a result of the relative velocities between transport vessel deck and the hanging load. Impacts loads occur randomly and are of very short duration. Due to the inherent uncertainties in the input parameters it is not considered feasible to calculate these loads accurately.”

Therefore, EDRMedeso has, in collaboration with FMC Technologies, Rambøll and DNV, developed an analysis procedure on how to perform a drop test according to the requirements in 2.7-3 in ANSYS. This will be explained briefly in this present article.

The concept is the same for physical and simulated drop test. Both methods follow the same procedure and both methods are accepted. However, several reasons make it advantageous to perform a simulated drop test instead of a physical test:

  • A much easier way to perform the test
  • Time efficient
  • No material waste
  • Reduces risk that is present during physical testing
  • Only one person involved ( see picture below for the number needed for physical testing)

As a result, simulating the drop test in ANSYS is environmental and cost friendly and is an easy way to drop test the PO unit.

Figure 1: Full model that needs to be drop tested

It is highly recommended to run the simulation as explicit dynamic simulation; hence, ANSYS LS-Dyna, ANSYS Autodyne or ANSYS Explicit STR (Autodyne solver) license is required.

Steps on how to set up a drop test in ANSYS:

  1. Simplify the model by keeping only the most vital parts
  2. Replace complex geometries
  3. Model the floor
  4. Define contacts and body interaction
  5. Mesh
  6. Solve
  7. Evaluate the results according to DNV acceptance criteria

Figure 2: Simplified model

Model set up

According to DNV 2.7-3 “The suspended PO Unit shall be so inclined that each of the bottom side and end members connected to the lowest corner forms an angle of not less than with the floor. However, the greatest height difference between the highest and lowest point of the underside of the unit corners need not be more than 400 mm. “

Figure 3: set up in Design modeler

Connections

Bonded body interactions are not supported for ANSYS LS-Dyna, so the bonded contacts are defined separately in this article. The frictional contact regions are shown in Figure 4 and Figure 5 and are defined as body-interactions.

Figure 4: Contact between dummies and bottom frame

Figure 5: Contacts between basket and floor

Boundary conditions

Apply an initial velocity equal to 1 m/s to the basket. Apply fixed support to the floor which is given rigid behavior. Apply standard earth gravity to all bodies.

Figure 6: boundary conditions

1. Evaluating Results

The design criterion according to DNV 2.7-3 reads; « No significant permanent damage shall occur. Cracks in welds and minor deformations may be repaired.” A typical acceptance criterion that has been agreed with DNV is that the portable offshore unit shall therefore rest on all corners after the test. This is demonstrated in Figure 7

Note: the post-processing part done in this article is done in LS-PrePost.

By selecting nodes on all four corners and one on the floor, the distance in z – direction can be visualized.

Figure 7: Position of corners

In addition, large volumes with plastic strain values above 2 % shall be evaluated, since this may represent significant permanent deformation.

Two animations can be seen below, where the first is a simulated drop test and the second is a physical drop test. Note that the animations do not represent the same model