Module 6: Steel Tension Members

Steel tension-member design is governed principally by NSCP 2015 Section 504 — Design of Members for Tension together with the connection provisions of Sections 510–511 where applicable.

A tension member normally requires at least two member-level checks:

1. Gross-section yielding — distributed yielding across the unreduced gross area.
2. Effective-net-section rupture — fracture through a reduced section near holes or connection discontinuities.

The connection can then introduce additional limit states such as block shear, fastener shear/bearing, weld rupture, local yielding, and eccentricity.

NSCP 2015 Chapter 5 adapts the AISC 14th Edition framework, which is based on ANSI/AISC 360-10. Within that basis, the two primary tension-member limit states use different reliability factors:

Compare required strength with the available strength for each limit state separately. Do not calculate both nominal strengths and then apply one common factor after selecting the smaller value.

Block shear and connection limit states have their own resistance/safety factors under the governing provisions and must be converted independently.

For steel plates and shapes, a potential rupture path may pass through staggered holes. Determine the net area using the NSCP steel net-section provisions, including the stagger correction where applicable, and examine all plausible critical paths rather than only the straight line across the fewest holes.

The simulation below uses the entered code-required hole deduction width $d_h$. That input should not be confused with a universal nominal bolt diameter because the required deduction depends on the governing hole provisions.

This steel procedure must not be transferred to timber.

When only part of a section is connected, stress cannot become uniform instantaneously across the full cross-section. The effective net area accounts for this shear-lag behavior. Connection length, eccentricity between connected and unconnected elements, and section type can influence $U$.

Longer, more distributed connections generally reduce shear lag, but the actual factor must come from the governing NSCP provision.

The effective net area is $A_e=UA_n$. Determine $U$ from the actual connection geometry and the adopted NSCP/AISC-based case; do not assign $U=1.0$ simply because the member is symmetric.

For connection cases where the code permits the centroid-offset form, the permitted relationship is

where $\bar{x}$ is the distance from the connection plane to the member centroid and $l$ is the connection length in the load direction. Other connection geometries use their own tabulated/case-specific $U$ provisions.

Use $A_e=UA_n$, but select $U$ from the connection configuration, not from member symmetry alone.

1. Establish which cross-sectional elements actually transmit force into the connection.
2. If the adopted provision permits the centroid-offset expression, determine the connection length in the load direction and the centroid-to-connection-plane offset before using $U=1-\bar{x}/l$.
3. If the geometry matches a section-specific/table case, use that case instead of forcing the centroid-offset expression onto it.
4. If all elements are directly connected, verify the provision that permits full effective participation rather than assuming $U=1.0$ automatically.
5. Re-evaluate $U$ when the connection length, connected leg/flange/web, bolt pattern, weld layout, or load-transfer plane changes.

The same member can therefore have different effective net areas under two different connection details.

Block shear combines shear failure along one or more planes parallel to the load with tensile failure across a transverse plane. It is especially important near bolted or welded connection ends.

A member that passes gross yielding and net-section rupture is not complete until the connection region is checked for block shear where applicable. Keep block shear independent from the member-level bar comparison because its areas and failure path belong to the connection region.

Very slender tension members may vibrate, sag, become difficult to handle, or experience unwanted secondary bending. Follow the NSCP user-note/recommendation applicable to tension-member slenderness rather than presenting a preferred slenderness value as a strength cutoff.

A complete tension-member solution should show: gross yielding, governing net-section rupture path, shear lag/effective net area, block shear, connection bolt/weld/connected-material limit states, and any practical slenderness/serviceability requirement.

Threaded regions have reduced effective tensile resistance. Where the NSCP/AISC fastener provisions express the nominal tensile stress of a threaded part as a fraction of $F_u$, apply that specified stress to the nominal area as directed by the provision.

Avoid casually stating that the physical tensile area is always exactly $0.75A_g$; that can be an equivalent simplification for a particular stress model but is not a universal geometric identity for threads.

Tie rods, hangers, bracing, suspended canopies, exposed trusses, and cable-supported secondary systems often make tension members visually prominent. Connection eccentricity, turnbuckles or adjustment, corrosion access, vibration, fire exposure, drainage, and erection tolerances should be resolved with the architectural concept.

A very slender exposed rod may satisfy strength but still be visually or dynamically unsuitable.

• Association of Structural Engineers of the Philippines (ASEP) — National Structural Code of the Philippines, C101-15, Volume I, Seventh Edition (2015) — Governing course reference for Section 504 tension-member provisions and Sections 510–511 connection requirements; consult the official licensed code for complete equations, tables, exceptions, and detailing provisions.