Shoulder Mechanics & Bench Press Surface Design

Broader scientific context

How do the shoulder, torso, bar, and bench interact?

Grip, scapular setup, torso position and bench contact shape the pressing task. Research shows how these variables affect bar movement, muscle activity and modeled shoulder loading.

The bench supports the posterior torso and helps determine shoulder clearance, effective height and resistance to sliding. Those contact conditions work with grip, bar path and torso position to shape the lift. Research can examine each part through movement, force and muscle-activity measurements.

Layer 01 · Anatomy

How the Shoulder’s Parts Work Together

The humeral head articulates with the glenoid at the glenohumeral joint. The scapula moves relative to the thorax and connects through the acromioclavicular joint to the clavicle; the clavicle connects to the sternum at the sternoclavicular joint. Rotator-cuff muscles contribute direction-specific control, while larger muscles generate and transmit pressing force.

Assess scapular movement in context

Clinical interpretation of scapular movement depends on the athlete, task, load and symptoms. Reviews and consensus statements recommend considering those factors together.[1][2]

AMM Research · Anatomical orientation

The shoulder chain and its support boundary

Posterior view of shoulder anatomy, emphasizing scapula on the thorax.

Posterior view

Scapula on the thorax

Both shoulder blades lie on the posterior rib cage. The socket-bearing scapula connects the upper arm to the shoulder girdle.

Anterior view of shoulder anatomy, emphasizing clavicle, sc and ac joints.

Anterior view

Clavicle, SC and AC joints

The sternoclavicular (SC) joint connects the clavicle to the sternum. The acromioclavicular (AC) joint links its lateral end to the scapula.

Shoulder detail of shoulder anatomy, emphasizing glenoid, humerus and cuff.

Shoulder detail

Glenoid, humerus and cuff

The humeral head meets the glenoid. Cuff muscles and tendons contribute to control around this articulation.

Follow the bony connection: sternum → SC joint → clavicle → AC joint → scapula → glenoid and humeral head.
Posterior support boundary

The bench meets the back of the torso. Contact and edge clearance belong to the interface record.

See the equipment contact example →
Vertical bar force

Force at the hands contributes to the external loading record. Its magnitude needs a defined measurement.

Mediolateral bar force

Force along the bar can alter calculated moments even when the visible path looks similar.

Follow the force record →
Illustrative anatomical views of the linked shoulder structures. Red highlights selected structures; the bench contacts the posterior torso.[1][2]

Layer 01 · Anatomy, continued

Why the Moving Socket Matters

The socket of the shoulder joint, the glenoid, is part of the shoulder blade. Wherever the scapula goes, the socket goes with it. When the shoulder blade glides around the rib cage, rotates or tilts, the socket turns, and the upper arm meets it at a different angle even if the arm has not moved relative to the torso.

That is the link between a bench and the joint itself. On a bench the shoulder blades lie between the rib cage and the pad, so whatever changes where and how they are supported can change how the socket sits under the arm.

In ordinary arm movement the two parts never work separately. Ludewig and colleagues fixed motion sensors to pins in the collarbone, shoulder blade and upper arm of 12 people without shoulder problems and recorded them raising and lowering the arm in three directions[9]. All four shoulder joints contributed substantial rotation: the collarbone rotated backward, the shoulder blade rotated upward and tilted back, and the upper arm rose and rotated in the socket. The study measured arm raising rather than pressing, but it establishes what the rest of this article relies on: the arm and the shoulder blade move as one system.

AMM Research · Anatomy · right shoulder

The shoulder blade carries the socket

Three views of the same joint: the socket on the shoulder blade, what happens to it when the shoulder blade moves, and what holds the front of the joint.

Close-up of a right shoulder. The socket of the shoulder joint, the glenoid, is highlighted on the shoulder blade; the upper arm bone is shown see-through in front of it.1Socket (glenoid)2Upper arm (humerus),
shown see-through
  1. Socket (glenoid)
  2. Upper arm (humerus), shown see-through

1 · The socket belongs to the shoulder blade. The glenoid, highlighted, is a shallow cup on the outer corner of the scapula, and the ball of the upper arm sits against it. Wherever the shoulder blade goes, the socket goes with it.

Two views of the same right shoulder from behind. Left: the shoulder blade is drawn back, close to the spine. Right: it has glided forward around the rib cage, farther from the spine, while the upper arm points the same way.Drawn back
retracted
Drawn forward
protracted
1Close to
the spine
2Socket faces
the side
3Farther from
the spine
4Socket turns
toward the front
  1. Close to the spine
  2. Socket faces the side
  3. Farther from the spine
  4. Socket turns toward the front

2 · Same arm, different socket. Seen from behind. On the left the shoulder blade sits close to the spine; on the right it has glided forward around the rib cage. The upper arm points the same way in both views, but the socket has turned toward the front, so the arm now sits further behind it.

Front of a right shoulder joint. The anterior band of the inferior glenohumeral ligament, at the front and bottom of the joint capsule, is highlighted.1Upper arm (humerus)2Front of the
joint capsule
  1. Upper arm (humerus)
  2. Front of the joint capsule

3 · Different demands on what holds the joint. Highlighted is the anterior band of the inferior glenohumeral ligament, at the front of the joint capsule. In the cadaver study described below[10], strain in this band rose as the shoulder blade was drawn forward. With the arm further behind the socket, the front of the capsule takes more strain, the rotator cuff works to keep the ball centered, and the shoulder-blade muscles and the bench determine where the socket is held.

Panels 1 and 3: stills from Glenohumeral Joint: Anatomy and Function, Muscle & Motion (Amit Gal Alon, Uriah Turkel). Panel 2: AMM illustration, showing the direction of the movement rather than its typical size. None of the panels is a measurement taken during bench pressing.

Three ways to describe the same arm position

Because the socket moves, “shoulder movement” can refer to three different measurements. Keeping them apart prevents the most common misreadings of bench-press research.

What is describedWhat it capturesWhat it cannot show on its own
Upper arm relative to the torsoThe combined movement of the whole shoulder complex, close to what a camera beside the bench recordsHow much happened at the joint and how much at the shoulder blade
Upper arm relative to the shoulder bladeGlenohumeral motion: the angle at the joint itself, with the socket as the referenceWhere the ball sits within the socket
Ball shifting within the socketHumeral-head translation, a separate, millimeter-scale motion usually measured with imagingAnything from bar path, video or arm angle alone

Two repetitions that look identical from the side can involve different arm-to-socket relationships if the shoulder blades sit differently. A visibly different bar path does not, by itself, show that the ball moved within the socket, and it says even less about whether any such movement was harmful. Studies that answer these questions measure segment motion, joint translation and external force separately.

Position changes what the passive structures do

Weiser and colleagues tested how shoulder-blade orientation affects the ligaments at the front of the joint[10]. They mounted five cadaver shoulders in the apprehension position, with the arm out to the side and rotated back, then simulated 10° and 20° of scapular protraction, the shoulder blade drawn forward around the rib cage, while pushing the ball of the humerus forward with set loads.

As protraction increased, strain rose in the anterior band of the inferior glenohumeral ligament, one of the main restraints at the front of the joint. Forward sliding of the ball decreased at the same time. Under a 15 N load it measured 6.3 mm with no protraction, 4.1 mm at 10° and 2.5 mm at 20°. The tightened ligament held the ball more firmly while carrying more strain.

These were cadaver joints in a fixed position rather than lifters under a bar, so the study does not measure what happens during a bench press. What carries over is the principle. The orientation of the shoulder blade changes the demand on the structures that hold the joint, even when the arm’s position relative to the torso is unchanged. It also shows why tissue demand and visible displacement are separate questions: a joint can move less while one of its restraints works harder.

Layer 02 · Kinematics

First Ask What Moved

Three distinct records

Name the motion before drawing a conclusion

Bar

Path, displacement and timing.

Shoulder girdle

Scapular translation, rotation and tilt.

Upper arm / torso

Humeral excursion and thoracic posture.

Record bar movement, shoulder-girdle motion and upper-arm/torso motion separately.

Kinematics describes position and motion. Motion capture, imaging and video-based methods track how the humerus, scapula, torso and bar move through the press.

Method

Motion capture · imaging · video-based kinematics

Output

Angles · translations · paths · timing

Inference limit

Movement is not a direct measurement of joint force, tissue stress, pain or injury.

Define the shoulder movement

Specify the direction and phase: retraction, protraction, rotation, tilt or translation. Assess that motion alongside control of the press.

Layer 03 · Kinetics and models

Then Ask What Force or Moment Was Estimated

Method

Instrumented bar · inverse dynamics · musculoskeletal modeling

Output

External forces · net joint moments · model-estimated reaction forces and muscle demand

Inference limit

Model outputs depend on assumptions; they are not in-vivo tissue measurements or clinical outcomes.

Read the participant and measurement protocol

Noteboom and colleagues tested 10 healthy, experienced strength athletes (nine men and one woman, with 6.7 ± 3.9 years of bench experience) across 21 combinations. Grip widths were 1.0, 1.5 and 2.0 times bi-acromial width; shoulder-abduction targets were 45°, 70° and 90°; scapular conditions were neutral, retracted and released. Participants completed three repetitions with a 16 kg bar. Motion capture and an instrumented bar supplied inputs to an OpenSim shoulder model.[3]

Noteboom et al. · 2024 · Table 2

Technique changes the modeled demand

  • 10 experienced lifters
  • 21 technique combinations
  • 16 kg bar
  • 3 repetitions per trial

One selected modeled muscle part, compared across three technique components. Every panel uses the same 0–1 scale; BAW means bi-acromial width.

Grip width

1.0 × BAW0.42

SD 0.20 relative activity

1.5 × BAW0.56

SD 0.17 relative activity

2.0 × BAW0.72

SD 0.13 relative activity

relative activity · bars show means; SD printed for each condition

Supraspinatus anterior · modeled peak activity, maximum = 1. Study [3].

Scapular pose

Neutral0.62

SD 0.17 relative activity

Retracted0.50

SD 0.17 relative activity

Released0.60

SD 0.16 relative activity

relative activity · bars show means; SD printed for each condition

Supraspinatus anterior · modeled peak activity, maximum = 1. Study [3].

Abduction target

45°0.49

SD 0.17 relative activity

70°0.58

SD 0.17 relative activity

90°0.65

SD 0.16 relative activity

relative activity · bars show means; SD printed for each condition

Supraspinatus anterior · modeled peak activity, maximum = 1. Study [3].
Modeled peak supraspinatus anterior activity, reported as means and SDs. Peaks were averaged within repetitions and trials, then across participants [3].
Measured

Body and bar kinematics; mediolateral (sideways) bar force, recorded by a strain-gauge barbell.

Assumed input: vertical force, set from the 16 kg bar weight and split equally between the hands.

Modeled

Selected glenohumeral and acromioclavicular reaction forces; selected muscle activity.

Not measured

In-vivo joint force, ligament loading, pain, injury, production bench-pad differences or heavy working sets.

Read the joint-force findings and model methods

Narrower grips and scapular retraction reduced selected acromioclavicular compression, glenohumeral posterior shear and modeled rotator-cuff activity during parts of the repetition. Mediolateral bar force materially affected the estimates. The released condition used a pool noodle under the spine, and released versus neutral conditions did not differ in modeled joint-reaction force. Twelve percent of trials were excluded for missing marker or lateral-force data, and ligaments were absent from the model.[3]

Technique changed selected model-estimated shoulder loads under the tested light-load conditions.

The Visible Bar Path Is Not the Whole Force Record

Force direction · measurement framework

A bar path does not contain the whole force record

Record the external forces

Vertical and mediolateral force components at the bar.

Combine with body geometry

Joint positions and moment arms through the repetition.

Interpret the calculated outcome

Net joint moments or model-estimated reaction forces, under stated assumptions.

Calculation framework combining external forces with body geometry [4][5].

In 35 strength-trained adults, Mausehund and colleagues found condition differences of up to 12% in load, up to 43% in shoulder net moments with wider grips and up to 26% in elbow moments as grip narrowed. Their central methodological point was that lateral bar forces materially change calculated joint moments.[4]

Related evidence: grip strategy and torso technique

Larsen and colleagues confirmed that grip affected both the weight lifted and the horizontal-force strategy during 1RM pressing in 14 experienced men.[5] Cudlip and colleagues added a torso-technique comparison: in 20 experienced men at 25%, 50% and 75% of self-reported 1RM, the powerlifting arch increased latissimus dorsi activity, while the standardized technique produced approximately 8% larger integrated shoulder moments.[6] The arch redistributed muscular activity and joint moments in this task.

Layer 04 · Muscle activity

Electrical Activity Is Direction-Specific

Anatomy context · Muscle roles

Several muscles contribute to the press

Anatomical press illustration identifying pectoralis major, anterior deltoid and triceps brachii, with shoulder horizontal adduction and elbow extension labels.
PECTORALIS MAJOR

Moves the upper arm

Helps bring the upper arm across the chest during the press.

ANTERIOR DELTOID

Assists shoulder motion

Contributes to shoulder motion; its role depends on the task and position.

TRICEPS BRACHII

Extends the elbow

Contributes to straightening the elbow during the upward press.

Anatomical orientation: colors identify the muscles. Muscle-role context: Rodríguez-Ridao et al. (2020). Select the image to enlarge.
Method

Surface or fine-wire electromyography

Output

Recorded electrical amplitude and timing for sampled muscles

Inference limit

EMG is not direct joint force, hypertrophy, pain or injury risk.

Wattanaprakornkul et al. · 2011 · EMG comparison

Pressing and rowing recruit the cuff differently

Bench press

Infraspinatus activity exceeded subscapularis activity.

Row

Subscapularis activity exceeded infraspinatus activity.

Direction-specific cuff recruitment across the tested loads [7].
Read the muscle-recording protocol and full explanation

Wattanaprakornkul and colleagues recorded nine shoulder muscle sites in 15 participants during bench press and row tasks at 20%, 50% and 70% of maximum load. Infraspinatus activity exceeded subscapularis activity during the bench press, while the pattern reversed during the row.[7] The cuff response was reciprocal and task-direction-specific, rather than one generic “stability” signal.

Layer 05 · Surface interaction

Where Contact With the Bench Enters the Mechanics

A bench establishes contact width, contour, friction, compliance, effective height and an edge around the posterior torso. These properties can affect the setup an athlete uses, resistance to sliding, rack relationship and the contact encountered near the scapular region. They are legitimate independent variables.

AMM Research · Visual 25Equipment photograph

Torso support and shoulder clearance

The Launch Pad® Gen II installed on an adjustable bench, viewed from the lumbar end, showing its contoured, arched back-support sections.

The Launch Pad® supports the upper back while its contoured edges create clearance around the shoulders and upper arms.

Advanced Muscle Mechanics product-image archive. Original image, unaltered.

Contact

Where and how much of the thorax and pelvis are supported under load?

Clearance

Which edge or contour approaches the upper torso through the repetition?

Constraint

How do friction and compression affect position or drift?

Response

Do measured kinematics, force, EMG or performance change?

The Athlete-to-Bench Interface covers the full research sequence. The acute study below examines pressing activation and bar movement.

Direct product evidence · interpreted at the measured layer

The Acute Launch Pad Study Changed the Support Condition

In the randomized crossover by Kidwell and colleagues, 10 resistance-trained men performed five repetitions at 70% of each participant’s established 1RM on a conventional flat bench and with the Launch Pad.[8] The Launch Pad condition produced higher bilateral pectoralis sEMG, mean and peak bar velocity and vertical displacement. Concentric power was 249 ± 103 W versus 271 ± 93 W, but the difference was not statistically significant (p = .071).

AMM Research · Measurement and meaning

Keep the measurement attached to its meaning

Standalone Vitruve velocity encoder device.

01 · Observation

Bar moves

Measurement
Position + time
Interpretation
Bar kinematics
Surface EMG electrodes on the pectoral region.

02 · Observation

Muscle is active

Measurement
Surface EMG
Interpretation
Electrical activity
A pressure-sensing sheet on a bench pad.

03 · Observation

Support changes

Measurement
Contact / pressure
Interpretation
Interface condition
Illustrative measurement categories. The acute Launch Pad study recorded muscle activity and bar movement [8].

Kidwell et al. · 2026 · Table 1

Six measured outcomes, with the full result in view

  • 10 trained men
  • 70% 1RM
  • 5 repetitions
  • Randomized crossover

Right pectoralis sEMG

Flat bench70.2

SD 24.8 % MVC

Launch Pad87.3

SD 20.1 % MVC

% MVC · bars show means; SD printed for each condition

p < .001 Study [8].

Left pectoralis sEMG

Flat bench68.3

SD 23.4 % MVC

Launch Pad86.8

SD 20.0 % MVC

% MVC · bars show means; SD printed for each condition

p < .001 Study [8].

Mean bar velocity

Flat bench0.40

SD 0.09 m/s

Launch Pad0.47

SD 0.09 m/s

m/s · bars show means; SD printed for each condition

p < .001 Study [8].

Peak bar velocity

Flat bench0.57

SD 0.11 m/s

Launch Pad0.66

SD 0.09 m/s

m/s · bars show means; SD printed for each condition

p < .001 Study [8].

Vertical bar displacement

Flat bench38

SD 7 cm

Launch Pad44

SD 4 cm

cm · bars show means; SD printed for each condition

p = .005 · bar travel, not shoulder-joint ROM Study [8].

Concentric power

Flat bench249

SD 103 W

Launch Pad271

SD 93 W

W · bars show means; SD printed for each condition

p = .071 · no statistically significant difference Study [8].
All six performance outcomes are shown. Bars start at zero, means and SDs retain the published units, and the power result remains nonsignificant. Pectoral activity and bar motion cannot establish a scapular or clinical mechanism. Study [8].
View the original compact numerical summary
Pectoralis sEMGRight: 70.2 ± 24.8 → 87.3 ± 20.1% MVCLeft: 68.3 ± 23.4 → 86.8 ± 20.0% MVCBoth p < .001
Bar velocityMean: 0.40 ± .09 → 0.47 ± .09 m/sPeak: 0.57 ± .11 → 0.66 ± .09 m/sBoth p < .001
Vertical displacement38 ± 7 → 44 ± 4 cmLinear-transducer bar displacement, not shoulder ROMp = .005
Concentric power249 ± 103 → 271 ± 93 WPoint estimate favored the pad conditionp = .071 · not significant
Measured

Pectoralis sEMG · bar velocity · vertical displacement · concentric power

Not measured

Scapular kinematics · shoulder-joint reaction force · rotator-cuff EMG · pain · injury

The tested support condition changed pectoralis activity, bar velocity and vertical bar displacement during the submaximal set.

Layer 06 · Clinical meaning

Shoulder Mechanics: Findings and Practical Implications

Grip, scapular setup, torso position and the support surface can change the mechanics of a press. Use those variables deliberately: establish a repeatable setup, check shoulder and upper-arm clearance through the intended range, and assess the athlete’s control and response under load.

For coaching, change one variable when possible and record the condition precisely. For research, match the conclusion to the method. For persistent pain, neurological symptoms, postoperative restrictions or return-to-training decisions, use appropriate clinical assessment rather than a mechanical inference.

Continue the mechanics sequenceScapular Motion During the Bench Press explores the motion terms in greater depth

References

  1. Ludewig PM, Reynolds JF. (2009). The association of scapular kinematics and glenohumeral joint pathologies. Journal of Orthopaedic & Sports Physical Therapy, 39(2), 90–104. doi:10.2519/jospt.2009.2808. ↩
  2. Kibler WB, Ludewig PM, McClure PW, Michener LA, Bak K, Sciascia AD. (2013). Clinical implications of scapular dyskinesis in shoulder injury: the 2013 consensus statement from the Scapular Summit. British Journal of Sports Medicine, 47, 877–885. doi:10.1136/bjsports-2013-092425. ↩
  3. Noteboom L, Belli I, Hoozemans MJM, Seth A, Veeger HEJ, van der Helm FCT. (2024). Effects of bench press technique variations on musculoskeletal shoulder loads and potential injury risk. Frontiers in Physiology, 15, 1393235. doi:10.3389/fphys.2024.1393235. ↩
  4. Mausehund L, Werkhausen A, Bartsch J, Krosshaug T. (2022). Understanding bench press biomechanics—The necessity of measuring lateral barbell forces. Journal of Strength and Conditioning Research, 36(10), 2685–2695. doi:10.1519/JSC.0000000000003948. ↩
  5. Larsen S, Gomo O, van den Tillaar R. (2021). A biomechanical analysis of wide, medium, and narrow grip width effects on kinematics, horizontal kinetics, and muscle activity on the sticking region in recreationally trained males during 1-RM bench pressing. Frontiers in Sports and Active Living, 2, 637066. doi:10.3389/fspor.2020.637066. ↩
  6. Cudlip AC, Maciukiewicz JM, Pinto BL, Dickerson CR. (2022). Upper extremity muscle activity and joint loading changes between the standard and powerlifting bench press techniques. Journal of Sports Sciences, 40(9), 1055–1063. doi:10.1080/02640414.2022.2046937. ↩
  7. Wattanaprakornkul D, Halaki M, Cathers I, Ginn KA. (2011). Direction-specific recruitment of rotator cuff muscles during bench press and row. Journal of Electromyography and Kinesiology, 21(6), 1041–1049. doi:10.1016/j.jelekin.2011.09.002. ↩
  8. Kidwell JA, Yamamoto T, Hetherton KJ, et al. (2026). Acute effects of thoracic-spinal elevation via a novel bench press pad on sEMG and barbell kinetics in resistance-trained males. International Journal of Exercise Science, 19(1), 1003. doi:10.70252/IJES2026103. ↩
  9. Ludewig PM, Phadke V, Braman JP, Hassett DR, Cieminski CJ, LaPrade RF. (2009). Motion of the shoulder complex during multiplanar humeral elevation. Journal of Bone and Joint Surgery (American), 91(2), 378–389. doi:10.2106/JBJS.G.01483. ↩
  10. Weiser WM, Lee TQ, McMaster WC, McMahon PJ. (1999). Effects of simulated scapular protraction on anterior glenohumeral stability. American Journal of Sports Medicine, 27(6), 801–805. doi:10.1177/03635465990270061901. ↩